Investigation of the Acoustic Response of a Confined

Portland State University Portland State University

PDXScholar PDXScholar

Dissertations and Theses Dissertations and Theses

Spring 7-17-2018

Investigation of the Acoustic Response of a Investigation of the Acoustic Response of a

Confined Mesoscopic Water Film Utilizing a Confined Mesoscopic Water Film Utilizing a

Combined Atomic Force Microscope and Shear Combined Atomic Force Microscope and Shear

Force Microscope Technique Force Microscope Technique

Monte Allen Kozell Portland State University

Follow this and additional works at httpspdxscholarlibrarypdxeduopen_access_etds

Part of the Nanoscience and Nanotechnology Commons and the Physics Commons

Let us know how access to this document benefits you

Recommended Citation Recommended Citation Kozell Monte Allen Investigation of the Acoustic Response of a Confined Mesoscopic Water Film Utilizing a Combined Atomic Force Microscope and Shear Force Microscope Technique (2018) Dissertations and Theses Paper 4451 httpsdoiorg1015760etd6335

This Thesis is brought to you for free and open access It has been accepted for inclusion in Dissertations and Theses by an authorized administrator of PDXScholar Please contact us if we can make this document more accessible pdxscholarpdxedu

Investigation of the Acoustic Response of a Confined Mesoscopic Water Film Utilizing a

Combined Atomic Force Microscope and Shear Force Microscope Technique

Monte Allen Kozell

A thesis submitted in partial fulfillment of the

requirements for the degree of

Master of Science

Physics

Thesis Committee

Andres La Rosa Chair

Peter Moeck

Erik Sagravenchez

Portland State University

Abstract

An atomic force microscopy beam-like cantilever is combined with an electrical tuning

fork to form a shear force probe that is capable of generating an acoustic response from the

mesoscopic water layer under ambient conditions while simultaneously monitoring force

applied in the normal direction and the electrical response of the tuning fork shear force

probe Two shear force probes were designed and fabricated A gallium ion beam was used

to deposit carbon as a probe material The carbon probe material was characterized using

energy dispersive x-ray spectroscopy and scanning transmission electron microscopy The

probes were experimentally validated by demonstrating the ability to generate and observe

acoustic response of the mesoscopic water layer

Acknowledgments

I like to thank the following people who have mentored assisted guided and supported

me while I worked on this masterrsquos thesis Graduate students at PSU such as Dan Lankow

Theo Brockman Mike Hopkins and others who Irsquom probably forgetting but only in name

Family who knew I could do it thank you Dr Andres La Rosa for being a knowledgeable

resource and welcoming adviser Irsquod like to thank my graduate committee members Erik

Sanchez Peter Moeck and once again Andres La Rosa My coworkers at Lam Research

Finally Irsquod like to thank Helena Chew thank you for your support pretty lady

Table of Contents

Abstracthelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip i

Acknowledgmentshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip ii

List of Tableshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip vii

List of Figureshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellipviii

Chapter 1 A Brief Introduction to Probing Microscopyhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 1

Section 11 Previous Scanning Probe Researchhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip1

Section 12 Proposed Cantilever Fabricationhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip6

Section 13 Cantilever amp Tip Fabrication Requirementshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip8

Section 14 Interaction Forces in AFMhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip8

Section 15 Capillary Forceshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 9

Section 16 Electrostatic Forceshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 10

Section 16 Van der Waals Forceshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 11

Chapter 2 Normal Force Modelshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 12

Section 21 Sphere-Plane (SP) Force Modelhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 12

Section 22 Introduction to a simple force-displacement curvehelliphelliphelliphelliphelliphelliphelliphelliphelliphellip14

Section 23 A Double jump-to-contacthelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 15

Section 24 Estimating the initial snap-to-contact pointhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 17

Section 25 Cantilever ldquoStiffnessrdquo Introducedhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 20

Section 26 Mechanical Amplificationhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 21

Section 27 Lateral Spring Constanthelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 27

Section 28 Laser Alignment and Interferencehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 28

Chapter 3 Cantilever and Tip Fabricationhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 29

Section 311 Method 1-Focused Ion Beam Milled Cantilever Fabricationhelliphelliphelliphellip 30

Section 312 Reflective Laser Padhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 33

Section 313 Tip Base and Tip Fabricationhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 33

Section 314 Electron and Gallium Ion Beam Deposited Carbon Tipshelliphelliphelliphelliphelliphellip 34

Section 32 Method 2-Commercial AFM Cantilever Adhered to Tuning Fork Tine hellip42

Chapter 4 Formatting data and sampling techniqueshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 48

Section 41 Force Distance Zero Linehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 48

Section 42 Normal Dimension Zero Force Pointhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 49

Section 43 Z-height Zero Pointhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip50

Section 44 Normal Force Scalinghelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 51

Section 45 Cantilever Deflection Amplitudehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 51

Section 46 Deflection Sensitivityhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 52

Section 47 AFM Scanner Calibrationhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 53

Section 48 Measurements with a Lock-In Amplifierhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 60

Section 49 Lock-in Amplifier Sampling Ratehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 61

Section 410 AFM Noise Levelhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 62

Chapter 5 Experimental Procedureshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 65

Section 51 List of Experimentshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip65

Section 52 Experiment 1 Directly Measuring Tuning Fork Oscillation Amplitudehellip 66

Section 53 Experiment 2 AFM Normal force curves to obtain Hamakerrsquos Constanthellip 73

Section 531 Experimental Force Curveshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 74

Section 532 Estimated Hamaker Constanthelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 75

Section 533 Probe Reliabilityhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 76

Section 534 STEM Analysis of IBD carbonhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip78

Section 54 Experiment 3 EDX characterization of EBD and IBD carbon materialshellip 79

Section 541 EDX Characterization of Electron and Ion deposited Carbonhelliphelliphelliphellip79

Section 542 Electron Beam Acceleration of 5k Voltshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 80

Section 543 Electron Beam Acceleration of 15k Voltshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 80

Section 544 EDX Resultshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 82

Section 55 Experiment 4 Surface interaction using FIB milled Cantileverhelliphelliphelliphelliphellip85

Section 551 Experimental designhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 85

Section 552 Measuring Normal Forcehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 85

Section 553 Measuring the Shear Forcehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 86

Section 554 Measuring the Acoustic Responsehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 87

Section 555 Substrate Preparationhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 87

Section 556 Performing a measurement with a Lock-in Amplifier (SR850)helliphelliphelliphellip 90

Section 557 XE-120 AFM considerationshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 90

Section 558 Experimental Resultshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 91

Section 559 Experimental Discussionhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 95

Section 5510 Experimental Limitationshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 98

Section 56 Shear force experiments using a Commercial AFM cantileverhelliphelliphelliphellip 99

Section 561 Experimental Designhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 99

Section 562 Measuring Normal Forcehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 100

Section 563 Measuring the Shear Forcehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 101

Section 564 Acoustic Amplitude and Lateral Displacement Voltage Measurements 101

Section 566 Measuring the Lateral Displacement Voltagehelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 102

Section 568 Performing a measurement with a Lock-in Amplifier (SR850)helliphelliphellip 106

Section 5610 Experimental Resultshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 107

Section 5611 A lsquoKinkrsquo in the Normal Force Curvehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip124

Section 5612 A lsquoStatic Clampingrsquo of the Tuning Forkhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip124

Section 5613 Estimated Tip Velocitieshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip125

Section 5614 Region A Complete Contact Regionhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 126

Section 5615 Region B The lsquoKnockingrsquo and Partial Contact Regionhelliphelliphelliphelliphelliphellip 126

Section 5616 Region C The 3rd Body Contact Regionhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 127

Section 5617 Region D The Mid-Range Interaction Regionhelliphelliphelliphelliphelliphelliphelliphelliphellip129

Section 5618 Region E The Long Range Non-Interaction Regionhelliphelliphelliphelliphelliphelliphellip 129

Section 5619 Plausible Explanations for the Observed lsquoBumprsquo in Region Chelliphelliphellip 129

Section 5620 Transition from Attractive to Repulsive Responsehelliphelliphelliphelliphelliphelliphelliphellip 131

Chapter 6 Future Experimentshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 133

Section 61 Future Force Curve Experimentshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 133

Section 62 Future Experiment 1 Subsurface void localizationhelliphelliphelliphelliphelliphelliphelliphellip133

Section 621 Proposalhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 133

Section 622 Implementationhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 134

Section 623 Sample Fabricationhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 134

Section 625 AFM Topographyhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 142

Section 616 AFM Normal Force Imagehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 142

Section 617 AFM Lateral Force Imagehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 143

Section 628 UFM Amplitude Imagehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 143

Section 629 UFM Phase Imagehelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 143

Section 63 Future Experiment 2 Defect Adhesionhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 145

Section 631 Defect Preventionhelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 145

Section 64 Surface Nano-Roughness Characterization using Acousticshelliphelliphelliphelliphellip 147

Section 643 Estimated Resultshelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphelliphellip 148

List of Tables

Table 1 Range of Force Interaction 9

Table 2 Estimated snap-to-contact point 19

Table 4 Varying Cantilever Length 25

Table 5 Varying Optical Path Length 26

Table 6 Equations for Spring Constants 32

Table 7 FIB Milled Cantilever Dimensions 32

Table 8 Material Properties of QTF and IBD Carbon 32

Table 9 FIB Milled Cantilever Spring Constants 32

Table 10 Summary of Electron-Beam and Ion-beam Tip Deposition Parameters 35

Table 11 Adhered Cantilever Dimensions 47

Table 12 Cantilever Material properties 47

Table 13 Adhered Cantilever Spring Constants 47

Table 14 Hamaker Constants for IBD Carbon Tip and Mica Surface 75

Table 15 Summary of Electron-Beam and Ion-beam thin film Deposition Parameters 81

Table 16 Energy Table for EDX analysis 81

Table 17 Summary of EDX characterization 82

Table 18 Tuning Fork Shear Velocity (Estimated) 125

Table 19 Summary of Feature Visibility 144

List of Figures

Figure 11 Schematic of AFM Operation 3

Figure 12 Schematic demonstrates mesoscopic layer-probe interaction 5

Figure 13 Illustrates an afm cantilever directly applied to a tuning fork to act as a force

sensor 7

Figure 15 Illustrates the formation of a liquid capillary bridge between a sample

substrate and an afm probe approximated as a sphere 10

Figure 21 The figure illustrates the typical force curve with adhesion hysteresis present

in the retraction curve 14

Figure 22 The figure shows an approach curve illustrating a double jump-to-contact 17

Figure 23 Schematic illustrating mechanical amplification 22

Figure 25 Illustrates how decreasing the mounting height of a cantilever in the z-axis

results in an increase in the optical path length 26

Figure 26 A typical normal force approach curve that illustrates laser interference 29

Figure 31 A schematic illustration showing the tuning fork with a beam cantilever that

was fabricated using a gallium Focused Ion Beam 35

Figure 31-1 Schematic illustrates the FIB milling of the cantilever onto the tuning fork

tine 36

Figure 31-2 Schematic illustrates the ion beam deposition of a reflective platinum pad

on the top of the beam cantilever 36

Figure 31-3 Schematic illustrates the top side of the FIB milled beam cantilever before

(A) ion beam platinum deposition of the laser reflective pad and (B) after 37

Figure 31-4 Schematic illustrates the ion beam deposition of a platinum base and then a

carbon tip 38

Figure 31-3 Schematic illustrates the bottom of the FIB milled beam cantilever (A)

before and (B) after the ion beam platinumcarbon deposition of the base and the tip

Figure 32 Schematic show the resulting FIB fabricated beam cantilever on tuning fork

tine 40

Figure 33 Shows the refined iteration a platinum tip base and the final carbon deposited

tip 41

Figure 34 Schematic illustrates afm cantilever transfer process 42

Figure 35 Illustrates the extraction of the afm cantilever inside the vacuum chamber of a

FIB-SEM FEI Helios 400s 44

Figure 36 Illustrates the welding of the afm cantilever onto a tuning fork tine inside the

vacuum chamber of a FIB-SEM FEI Helios 400s 45

Figure 37 Shows the actual tuning fork tines with an afm cantilever attached with

electron beam platinum welds 47

Figure 41 A schematic representation of the approach and retraction curve 50

Figure 42 High Voltage Z-Calibration Profile 55

Figure 43 Low Voltage Z-Calibration Profile 56

Figure 44 The STEM image shows the interface between iridium and bare silicon z-axis

calibration standard 57

Figure 45 A high resolution STEM image of the low-z lsquogluersquo perturbation 58

Figure 46 High Voltage X-Y Calibration Profile 59

Figure 47 Shows z-axis afm noise 63

Figure 48 Shows afm normal force noise 64

Figure 51 Illustrates an afm cantilever in direct contact with a stationary tuning fork 67

Figure 52 Schematic illustrates experiment setup for measuring tuning fork

displacement 68

Figure 53 A laser deflection voltage to displacement curve 70

Figure 54 Demonstrates relation between tuning fork displacements and measured

electrical response 72

Figure 55 This figure shows the observed normal force with the zero force and zero

distance points 75

Figure 55-1 Left SEM image is before and right is after acquiring 10 normal force

curves without a shear force applied 77

Figure 55-2 Image shows an example of probe wear after many (10rsquos to 100s) normal

force approach curves while a shear oscillation is applied 77

Figure 55-3 A bright-field STEM image showing the nano-structure of IBD carbon 78

Figure 56 Shows the EDX spectra for IBD carbon (Top) and EDB carbon (Bottom) 84

Figure 57 Illustrates the tuning fork oscillating motion relative to the sample surface 86

Figure 58 Experiment Schematic illustrating the normal force shear force and acoustic

interaction measurements 89

Figure 59 Experimental results for Trial 1 observing normal force shear force and

acoustic signal 92

acoustic signal 93

acoustic signal 94

Figure 512 Illustrates a comparison of the relative positioning of the acoustic response

obtained in current experiment and the double snap-to-contact positioning normal

force curve obtained in a verification experiment 97

Figure 513 Illustrates the tuning fork oscillation direction relative to the sample surface

(Image 1) 100

Figure 514 This schematic shows the basics of the experiment where normal force

tuning fork amplitude and acoustic amplitude are measured 104

Figure 515 The schematic shows the basics of the experiment where normal force

tuning fork amplitude and lateral displacement voltages are measured 105

Figure 516 Data-Set1 A series of normal Force vs Z-Displacement approach curves

plotted as tuning fork drive voltage is increased from 50mV to 400mV 108

Figure 517 Data-Set1 The series of Normal Force vs Z-Displacement approach curves

plotted overlapping as tuning fork drive voltage is increased from 50mV to 400mV

Figure 518 Data-Set1 A series of Tuning Fork Amplitude vs Z-Displacement

approach curves plotted as tuning fork drive voltage is increased from 50mV to

400mV 110

Figure 519 Data-Set1 The series of Tuning Fork Amplitude vs Z-Displacement

approach curves are overlapped to illustrate the general trends of increased tuning fork

amplitude damping 111

Figure 520 Data-Set1 A series of Acoustic Amplitude vs Z-Displacement approach

curves plotted as tuning fork drive voltage is increased from 50mV to 400mV 112

Figure 521 A low amplitude tuning fork approach curve with distinct interaction regions

labeled 113

Figure 522 A high amplitude tuning fork approach curve with distinct interaction

regions labeled 114

Figure 524 Data-Set2 The series of Force vs Z-Displacement approach curves plotted

overlapping as tuning fork drive voltage is increased from 14mV to 350mV 117

350mV 118

approach curves are overlapped to illustrate general trends of increased tuning fork

Figure 527 Data-Set2 A series of Lateral Displacement vs Z-Displacement approach

Figure 528 Data-Set2 The series of Lateral Displacement vs Z-Displacement approach

curves are overlapped to illustrate the increase in lateral displacement dampening as

the tuning fork drive voltage is increased 121

labeled 122

regions labeled 123

Figure 531 Illustrates the observed lsquokinkrsquo in the normal force curve and the static

lsquoclampingrsquo effect for the tuning fork amplitude 124

Figure 61 Demonstrates electron and ion beam fabrication of subsurface features 135

Figure 62 This illustrates the ion beam deposited carbon capping of the fabricated box

outlines 136

Figure 63 Shows a schematic outline of the fabricated features that have been buried

under an ion beam deposited carbon layer 137

Figure 64 Image illustrates a measurement of the carbon cap height of approximately

500nm 137

Figure 65 A 3D topographic afm image of the scanned features 138

under a ion beam deposited carbon layer 140

Figure 68 Schematic illustrating using an acoustic sensor-transducer to apply a force to

the surface of a wafer with surface debris present 146

Figure 69 (A) Illustrates an acoustic wave is incident upon a smooth surface that

exhibits specular reflection 148

Chapter 1 A Brief Introduction to Probing Microscopy

Surfaces and material properties have been investigated for much of the 20th

century The development of the scanning electron microscope the transmission electron

microscope the maturity of energy dispersive x-ray spectroscopy crystallography the

invention of the scanning probe microscope in its many variances and many other

microscopy techniques have propelled the current technological revolution that we

currently enjoy It can be stated that age of nanotechnology has arrived however the field

of acoustics has hardly progressed much towards being a truly nanoscale discipline Nano-

acoustic interactions as a form of sample manipulation or for sample characterization has

the potential to evolve into a powerful tool to characterize and manipulate biological

surfaces and investigate subsurface material properties To develop such tool we will

investigate the fundamentals of generating and monitoring nano-acoustics of surface

interactions

Section 11 Previous Scanning Probe Research

The field of shear force microscopy is still a relatively new and progressing field

In shear force microscopy a nanometer sized probe is adhered to a tuning fork oscillator

and electrically driven to oscillate at a free amplitude resonance frequency while the probe

is allowed to interact with a sample substrate The mechanical or electrical response of the

tuning fork is typically monitored as a control parameter allowing the ability to maintain

a fixed set point of height or force between a probe and substrate Typical benefits of such

interaction are that the probe can interact with the substrate at a fixed height above the

substrate operating in a shear motion removing a direct probe-sample interaction typical

of tapping mode contact mode and intermittent contact mode of atomic force microscopy

(AFM) By removing a direct mechanical interaction in theory a probe will avoid

mechanically impacting a sample surface extending probe lifespan and avoiding sample

damage

To measure relative separation distance between a sample and probe the shear

force microscope employs tunneling current of electrons between a probe and the sample

surface Researchers label an arbitrary tunneling current as the sample surface usually a

current of 1nA used to designate the probe-surface contact1 The tunneling distance

measurement method limits samples to clean and conductive samples

AFM by contrast is directly able to mechanically measure and detect a probe-

sample interaction using a flexible beam cantilever with a probe fabricated on the end Both

insulating and conducting samples can be used since the afm method is based upon a

mechanical interaction An afm can achieve sub-nanometer height resolution by using a

laser reflected off of an afm cantilever and onto a position sensitive photodiode (pspd) As

the afm probe is allowed to interact with the sample the cantilever and subsequently the

reflected laser are proportionately deflected Figure 11 below gives a schematic

representation of afm operation Considering that the afm feedback is cantilever based it

can easily measure many properties of a sample-probe interactions such as normal force

lateral force and an amplitude-phase response of a vertically oscillating cantilever A direct

mechanical observation of the probe-surface interaction is a fundamental capability that

shear force microscopy lacks and with which the shear force investigation of a wider class

of samples would be possible while maintaining the ability to correlate shear force results

with probe-sample separation distance

Figure 11 Schematic of AFM Operation Image A Illustrates the standard scanning head configuration A

cantilever is mounted on a z-piezoelectric scanner with a laser reflected off or the cantilever and onto a

position sensitive photodiode A force applied to the cantilever will result in a deflection of the laser spot

across the pspd Control electronics not shown operate a feedback loop to control the force applied to the

cantilever tip while scanning with the x-y scanners Image B illustrates the laser reflecting off of the

cantilever onto a pspd detector Notice how the pspd is divided into quads a b c d Vertical deflections are

defined as A-B with A = (a+b) B = (c+d) and lateral deflections defined as C-D with C = (a+c) D = (b+d)

A pspd gives an electrical signal proportional to measured light intensity and the relative positioning on the

pspd Image A amp B from PJ Eaton and P West Atomic force microscopy Oxford University Press Oxford

2014 pp23-242

The nature of the shear force interaction is a source of controversy in the field of

shear force microscopy Traditionally there have been two interpretations of the shear

interaction The first interpretation is that the shear force interaction and tuning fork

dampening is driven by a probe lsquoknockingrsquo on the sample surface before solid mechanical

contact is established Ulbrich3 Ulbrich et al and others 4 find that the degree of tuning

fork dampening is directly dependent upon probe radius and the relative angle between the

oscillatory motion of the probe and the sample surface Ulbrich also demonstrates that there

exists a direct relation between tuning fork drive amplitude and the distance over which

the tuning fork is dampened ie dampening occurs over increasingly larger distances as

shear amplitude is increased It is noted that Ulbrich operates the tuning fork at large shear

amplitudes of approximately 6-30nm In a second interpretation Karrai and others5 while

operating at sub-nanometer tuning fork drive amplitudes1 are not able to observe the

lsquoknockingrsquo between a probe and sample surface as observed by Ulbrich Karrai is able to

fit his observations using a standard driven harmonic oscillator while Ulbrich fits his

observations using a non-linear model

Acoustics in shear force microscopy is a new and relatively unexplored field In

recent publications6 there have been attempts to complement shear force microscopy by

the inclusion of the acoustic observation of the tuning fork probe and the 3rd body medium

interactions It has been observed that the probe-water interaction exhibits a dampening

effect on tuning fork amplitude and acoustic observations are direct mechanical

observations of probe-water interactions The figure 12 below illustrates the tuning fork

probe and sample surface interaction occurring under ambient conditions The generation

of an acoustic response is illustrated in figure 12-C

Figure 12 Schematic demonstrates mesoscopic layer-probe interaction Image A shows the typical NSOM

experimental setup A probe is adhered to a quartz tuning fork (QTF) and the probe is allowed to interact

with a sample surface where a mesoscopic water layer is present The separation distance is maintained by

measuring the electrical response of the QTF as the probe-mesoscopic layer interact As the interaction

occurs the electrical response of the tuning fork changes Image B shows the probe-sample interaction

mediated through a mesoscopic water layer The probe is stationary in this image Image C shows a probe

that is driven parallel to a sample surface The probe-mesoscopic interaction generates an acoustic signal as

depicted as radiating waves which travel both through the open air and through the sample Acoustic response

is detected in the near field regime

It is the purpose of this thesis to merge complementary aspects of the two fields of

microscopy afm and shear force microscopy by fabricating and characterizing a tuning

fork with a flexible afm beam cantilever adhered onto a tuning fork tine This will allow

the direct ability to mechanically observe the true sample separation height while

maintaining the ability to investigate the dampening nature of the shear interaction of a

probe and the mesoscopic water layer The inclusion of an acoustic sensor will allow the

direct ability to observe viscous andor elastic interactions Previous work has allowed the

implementation of shear motion in conjunction with normal force measurements7 with the

use of a piezoceramic mounted beneath the sample substrate In such work lateral force

measurements were made with the afm however acoustics signals were not observed

Section 12 Proposed Cantilever Fabrication

It was decided to incorporate a force sensitive cantilever onto a tuning fork as one modular

apparatus because there are instances in which it may not be feasible to piezoelectrically

vibrate a large sample ie full wafer analysis in situ biological experiments Incorporating

the cantilever furnished the option to place an acoustic sensor beneath the sample substrate

reducing the acoustic losses from an otherwise increased acoustic travel distance

We have built two separate apparatuses the first of which is an afm cantilever

directly adhered to a tuning fork This apparatus is seen in the figure 13 below The second

apparatus is a tuning fork which has a beam cantilever directly milled into a tine on the

tuning fork using the focused ion beam (FIB) milling process This apparatus is seen in

figure 14 below Ideal cantilevers will be force sensitive and used to acoustically excite

the mesoscopic region The two apparatuses will have a carbon nanoprobe acting as a afm

tipprobe which is ion beam deposited (IBD) onto the cantilever The use of a custom

probe allows the fabrication of a geometrically repeatable devices It has been previously

demonstrated that tip geometry plays a crucial role in tip-sample interaction force 8

therefore it greatly affords flexibility in experimental design to have the ability to tune tip

geometry

Cantilever Configuration 1

Figure 13 Illustrates an afm cantilever directly applied to a tuning fork to act as a force sensor (Image A)

Image B is a close up of the cantilever Fabrication will be described in later sections

Figure 14 Illustrates a cantilever that was focused ion beam milled into a tuning fork to act as a force sensor

(Image A) Image B is a close up that shows the milled cantilever and the ion beam deposited tip Fabrication

will be described in later sections

Section 13 Cantilever amp Tip Fabrication Requirements

To fabricate an afm cantilever that is able to investigate normal force interactions

a few key design parameters must be conceptually investigated The primary

considerations for probe design are the mechanical advantage of the system which is

directly related to cantilever sensitivity cantilever geometries such as cantilever length

width and thickness which are directly related to the stiffness for the cantilever and the

tip probe material and tip probe radius which are related to normal forces which dictate

whether the probe will comply with the sample surface upon interaction Each parameter

will be discussed with the goal of offering insight into how changes in each parameter

affect the ability of the afm cantilever to measure normal force and generate an acoustic

signal

Section 14 Interaction Forces in AFM

Before we begin to describe the process of probe fabrication it is useful to discuss

the interaction forces that we expect to encounter during our experiment There are many

probe-surface interaction mechanisms Table 1 below describes many of the interaction

mechanisms found in probing microscopy with the approximate interaction strength and

the effective range of the interaction labeled It is clear that there exists significant overlap

between many interaction mechanisms and interaction strengths It is difficult to isolate

one particular force as being the dominant factor as a function of distance This contributes

greatly to the uncertainty in interpreting force measurements It is often left up to the

experimenter to use best judgement and a scrupulously designed experiment to estimate

the forces present and to estimate the magnitude of each force We will do the same while

acknowledging the inadequacies of such method The main forces present in our

experiment are expected to be Van der Waals dipole forces possibly surface charges with

electrostatic interactions and the capillary forces from the formation of a mesoscopic layer

Table 1 Range of Interaction

Type of Force Dependence of energy on

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

London Dispersion 1d6 05 to 5

H-bonding 1d3 05 to 3

Dipoles 1d3 05 to 3

Electrostatic e-d 10s to 100s

Van der Waals 1d 5 to 10

Solvation e-d lt5

Hydrophobic e-d 10s to 100s

Ref PA Maurice Environmental surfaces and interfaces from the nanoscale to the global scale (Wiley

Hoboken 2009 P 94)89

Section 15 Capillary Forces

The main purpose of our experiment is to generate an acoustic signal from the

interaction with the mesoscopic water layer We will be conducting our experiments in

high humidities conditions at 60+ At high humidities there will be a mesoscopic water

layer present on both the afm probe and sample surface10 with assumption that both

surfaces are hydrophilic To gain an appreciation of the capillary interaction a typical

schematic is shown illustrating the formation of a capillary bridge between an afm probe

and a sample surface in the figure 15 Figure 15 is typically used for capillary force

modeling using the Derjaguin Muller Toporov (DMT) model This model will not be used

in our work due to the scale of our experiment but the schematic is mentioned here to

illustrate the capillary formation between a probe and a hydrophilic surface

Figure 15 Illustrates the formation of a liquid capillary bridge between a sample substrate and an afm probe

approximated as a sphere Critical dimensions are labeled Figure reproduced by following M Rosoff Nano-

Surface chemistry (Dekker New York 2002) P 23 11

Section 16 Electrostatic Forces

Electrostatic forces are observed in the presence of charge on the afm probe andor

the sample substrate These forces can be either attractive or repulsive depending upon the

magnitude of charge present In our experiment our sample substrates are SiO2 and mica

The SiO2 is insulative in nature and does not inherently carry a net charge Mica has been

shown to contain K+ ions in the presence of a mesoscopic water layer10 Additionally we

will be producing a gallium ion beam deposited (Ga-IBD) carbon tip as an afm probe

Through EDX we find that the (Ga-IBD) carbon probe contains 25 Ga atoms in

composition which is in agreement with other findings 12 Therefore the possibility of free

Ga ions remaining in the probe is present Having a sample surface which can contain free

K+ ions and an afm probe which can contain Ga+ ions there exists the possibility of ion-

ion repulsion

Section 16 Van der Waals Forces

Van der Waals forces are representative of induced dipole and quadrupole

interactions Van der Waals forces arise from quantum mechanical distribution of electrons

and are generally non-localized in nature Van der Waals forces are long range attractive

and are said to be a weak interaction force Most direct measurements of Van der Waals

forces are performed in high vacuum systems where other interaction mechanisms are

reduced however the Van der Waals force is present under ambient conditions as well9

Chapter 2 Normal Force Models

To properly model the measured normal force it is necessary to choose an

appropriate model that fits the experimental geometries such as probe radius interaction

distances and environmental conditions involved Our experiments will be performed from

the mechanical contact point with the sample surface up to a distance of about 100nm

above a sample surface The experiment will be performed under ambient conditions with

enhanced humidity to promote the formation of a mesoscopic water layer In our

experiments we used an afm tip of approximately 100-250 nm in radius Therefore it is

useful to investigate a model that is able to represent normal forces in both the short range

and the long-range regions while approximating in ambient conditions Such ldquocompleterdquo

force models do not exist therefore we will attempt of model only parts of the interaction

force curve where models are well defined and experimental conditions match the

experiment domain

Section 21 Sphere-Plane (SP) Force Model

To model the long-range interaction forces (Van der Waals) between the probe and

substrate we will consider the Sphere-Plane (SP) model which was derived by Hamaker

in 193713 Hamaker was able to obtain a relation that represents a geometrically simplified

approximation of an afm tip-surface interaction with the afm tip being approximated as a

sphere and the sample surface as a flat plane The force Fsp relation for the sphere-plane

model is given as 13

Fsp = -2AR33z2(z+2R)2 (EQ 21)

where A is the Hamaker constant with units of energy R is the probe radius and Z is the

distance between the spherical tip and the planar surface under static conditions The

Hamaker constant is an energy scalar that is used to represent the molecule to molecule

interactions Hamaker shows that A can be expanded to represent n number of materials

allowing for the modeling of complex multilayer interfaces such as the long range (non-

contact region) afm force interactions between a tip a mesoscopic layer and a lower

substrate The SP model is not valid for contact regions since it cannot model mechanical

contact and as such it should not be applied within contact regions

To use the SP model to correctly fit experimental data we will need to directly

measure the Hamaker constant A for the specific experimental materials and atmospheric

conditions used The Hamaker constant is dependent upon tip material sample substrate

material and the specific humidity conditions Different combinations of tip and substrate

materials will have different Hamaker constants Critics of the Hamaker constant point out

that it depends highly upon experimental conditions such as a change in humidity In

Hamakerrsquos landmark paper of 1937 he describes the Hamaker constant as a fitting

parameter that can be used to describe the attractive force component between two objects

separated by an arbitrary distance He describes the Hamaker parameter as being able to

represent an object that is a composite of multiple layers such as a substrate with a

mesoscopic film present Hamaker gives an example of two different composition particles

embedded in a third material a fluid giving the Hamaker constant in an expanded form

A = π2 q1q2λ12 + q02λ00 - q0q1λ01 - q0q2λ02 (EQ 22)

With qi being atom densities and the λi being the London-van der Waals constants for the

pairs of atoms indicated by the indicie13 This expansion of the Hamaker constant is used

to justify the validity for applying the Hamaker force model in multi-layered interactions

A direct measurement of the Hamaker constant and experimental fitting with the SP model

will be performed in chapter 5 for our experimental conditions

Section 22 Introduction to a simple force-displacement curve

A force-displacement curve is obtained as a cantilever is slowly lowered to interact

with a sample substrate while the normal force (vertical cantilever deflection) is monitored

and recorded A typical afm force-displacement curve is shown below The relative

cantilever bending motion is illustrated

Figure 21 The figure illustrates the typical force curve with adhesion hysteresis present in the retraction

curve Notice that attractive forces result in a cantilever bending towards the sample substrate while repulsive

forces result in a bending away from the surface Plot is taken from page 65 Atomic Force Microscopy by

Peter Eaton Paul West15

Notice that as the afm probe approaches the sample surface the attractive van der

Waals forces pull the tip into contact with the sample substrate This is known as a snap-

to-contact point and is labeled as point B in the figure above A cantilever acts as a spring

as force is applied the cantilever exerts a restoring force maintaining an equilibrium at

large tip-to-sample distances At point B the cantilevers restoring force is overcome and

the tip snaps into mechanical contact with the sample substrate This effect in known as

snap-to-contact and it prevents the probe from fully sampling the attractive forces as a

function of tip-to-sample separation distance One of the chief experimental considerations

of our experiment is the ability to characterize the mesoscopic water layer This layer exists

close to the sample surface where forces are strong and attractive therefore it is decided

that our cantilever should be ldquostiffrdquo so as to avoid a snap-to-contact Our definition of a

ldquostiffrdquo cantilever is a cantilever that has a large spring constant in the vertical direction

Parameters that define a cantilevers spring constant will discussed in a later section and a

quantitative estimate of a stiff spring constant will be provided

Section 23 A Double jump-to-contact

The previous section described a force curve that was representative of a force-

displacement interaction involving solely Van der Waals forces Under ambient conditions

there exists a mesoscopic water layer present upon hydrophilic surfaces As an afm probe

performs an approach curve onto a hydrophilic surface in high humidity conditions

(Humidity gt 60) it must pass through two distinct boundaries initially it must pass

through the mesoscopic water layer and secondly it will interact with the underlying

substrate SiO2 or Mica in our experiments Therefore it will experience two distinct and

observable snap-to-contacts16 An example of a double snap-to-contact is shown below in

figure 22 In the figure below the initial snap-to-contact is labeled as A This corresponds

to the tip-mesoscopic layer interaction The second snap-to-contact is labeled as B this

represents the tip-sample interaction Notice that the force at point B is not zero even

though the tip is in direct contact with the sample surface The zero-force point is located

approximately 6nm to the left At point B the cantilever can be thought of as being bent

downward into mechanical contact with the sample surface with the source of attraction

provided by Van der Waals and capillary forces As the cantilever approaches further

toward the substrate a repulsive electron-electron force (Pauli Exclusion Principle)

counteracts the attractive forces and a force equilibrium is temporarily achieved at the zero-

force point This zero-force point is identified as the intersection of a dotted horizontal line

and a solid vertical line in figure 22

Figure 22 The figure shows an approach curve illustrating a double jump-to-contact with the initial jump-

to-contact labeled as point A and the subsequent jump-to-contact labeled as point B At point A the interface

is a transition from air to water At point B the interface is water to mica The lower plot has the relative

positioning and observed forces at each interface This approach curve was obtained with humidity of 638

A non-contact cantilever of spring constant ~40Nm was used

Section 24 Estimating the initial snap-to-contact point

Predicting where a snap-to-contact point will occur is handled by Soma Das Sreeram and

Raychaudhuri17 It is found that the snap-to-contact position is related to the tip radius the

cantilever spring constant in the normal direction and the material interaction parameter

the Hamaker constant The relation has the form

Hj3 = 2724(HRtKc) (EQ 23)

with Hj being the jump into contact position H being the Hamaker interaction constant Rt

is the radius of the tip and Kc is the cantilever spring constant in the normal direction This

relation assumes only van der Waals interaction and a large radius tip is used To get an

estimate the snap-to-contact position of a Si3N4 cantilever on a mica surface Somas Das

et al used a tip radius Rt = 35nm cantilever spring constant Kc 01nNnm and a Hamaker

H= 064x10^-19J A jump-into-contact position of 29nm was found A general trend that

was observed was that the snap-to-contact position is both dependent upon the tip radius

and the tip cantilever spring constant The Hamaker interaction parameter acts as a scalar

and is highly dependent upon humidity Uncertainty in the snap-to-contact point is highly

dependent upon the zero-force point and the relative positioning of the sample surface the

zero-distance point

Using equation 22 and solving for the Hamaker constant the experimental

observation of a force curve can be used to obtain the Hamaker constant between two

known surfaces This allows the modeling of the attractive forces before the probe directly

contacts these surfaces In chapter 5 we will use this approach to obtain an estimated

Hamaker constant between mica and an ion beam deposited IBD carbon probe

To fabricate a cantilever that is able to probe down to the surface contact point

without a snap-to-contact point we need an estimate of the required spring constant for

such cantilever Using equation 22 we will try to estimate the snap-to-contact point for a

standard non-contact Si3N4 afm cantilever on a mica surface with a 100nm radius and a

spring constant of varying values We will use the same Hamaker constant as Somas Das

H= 064x10^-19J since the materials are the same The results are displayed in the table 2

below It is evident that the initial snap-to-contact will not occur until the tip is within a

sub nanometer of the sample surface given that the spring constant is sufficiently large

Table 2 Estimated snap-to-contact point

Spring Constant K Snap-to-contact point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Parameters H=064x10^-19J Surface=Mica Tip=Si3N4 Rt=100nm

Since we are using a custom tip which we do not yet know the Hamaker constant

this is only a guide for suggested cantilever spring constant values Hamaker constants can

vary from low values 10-21J to larger values 10-18J Using these Hamaker values with the

above relation (EQ 22) we can refine our estimate of the snap-to-contact point in an

attempt to predict a range of cantilever spring constants that are sufficient to prevent a

snap-to-contact for the probe we are designing

As shown in the table below it is observed that at low Hamaker constants ~10-21J

the interaction energy between the tip and sample is low therefore softer spring constant

cantilevers may be used However with large Hamaker constants gt10-18J the snap-to-

contact point occurs 10 times further away ie there is a larger interaction energy between

the tip and sample In our experiment there will be a capillary force present which is an

attractive force in addition to the attractive Van der Waals forces These forces are additive

in nature In order to prevent a premature snap-to contact by these two attractors it is

decided that a cantilever with a spring constant of a minimum of K = 50 Nm will be used

Spring Constant (K)

with Low Hamaker 10-

Snap-to-contact point Spring Constant (K)

with High Hamaker 10-

Snap-to-contact point

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

Parameters Surface=Mica Tip=Si3N4 Rt=100nm

It should be noted that this relation for predicting jump-to-contact point is only valid for

the first jump-to-contact point and cannot be used for the second jump to contact point

since the relation only accounts for Van der Waals forces and excludes complex capillary

effects that are present once the tip has entered the mesoscopic region

Section 25 Cantilever ldquoStiffnessrdquo Introduced

In the previous section it was illustrated that in order to fully characterize the mesoscopic

layer a ldquostiffrdquo cantilever is needed A cantilever spring constant is given as a functions of

cantilever material and geometries The vertical spring constant of an afm cantilever is

given as

Kver = wE4( t L)3 (EQ 24)

with Kver being the vertical spring constant of the cantilever w being the cantilever width

E being the Youngrsquos modulus of the cantilever t being the cantilever thickness and L being

the cantilever length15 We will be fabricating our cantilevers out of silicon and quartz so

the Youngrsquos modulus will be a fixed early on We are left with cantilever geometry as our

tuning parameters to fabricate a cantilever with a desired stiffness

Section 26 Mechanical Amplification

In narrowing our cantilever geometries down we will consider the role of cantilever

length on cantilever sensitivity These two properties are linked by a process known as

mechanical amplification The XE-120 afm used in our experiment utilizes an optical

mechanical amplification to increase the pspd sensitivity to a cantilever displacement

An ideal mechanical lever is a lever that ideally never flexes and only bends thus

an afm cantilever being of crystalline material is a close approximation The mechanical

advantage of a lever is defined as the ratio of two lengths that pivot around some central

point On an afm these two lengths correspond to the cantilever length and the optical path

that the laser takes once reflected from the cantilever to the pspd This is shown in the

figures 23 24 below In the context of an afm system mechanical advantage has the

following form

Mechanical Advantage = Optical Path Length Cantilever Length

(MA) = (L1L2)

Figure 23 gives two examples of mechanical advantage with one apparatus producing no

mechanical advantage and the other apparatus producing a gain of 3 Such gains are

achieved by altering the ratio of path lengths In the figure 24 it is illustrated how an afm

optical path and the cantilever constitute a system that exhibits mechanical amplification

Figure 23 Schematic illustrating mechanical amplification Top Shows an example of mechanical

amplification with a 11 ratio both lengths are equal Bottom Shows that mechanical amplification can be

modified by changing the relative lengths of each side Here L1 is increased to 15 and L2 is decreased to

05 The mechanical advantage is increased by a factor of 3

Figure 24 Demonstrates optical path

length and mechanical amplification of

an afm scanning head (Top) Shows the

standard afm geometry (Below) Shows

the optical path (red dots) and the

cantilever length (green line) Notice

how both lengths pivot around the afm

tip As the cantilever is lowered while in

contact with the substrate the cantilever

displacement distance is amplified by the

mechanical advantage producing a

larger displacement of the laser spot on

the pspd Source of image Park AFM XE-120

user manual page 29 18

By applying the principle of

mechanical advantage to the afm

cantilever and the optical path

length of the laser one can gain

insight into choosing a cantilever

length Consider the standard afm

tapping mode cantilever which

typically have a length of about

125um The optical path length of

the XE-120 afm is estimated to be between 40-50 cm with an average 45 cm Using these

geometries the mechanical advantage of the tapping mode cantilever is estimated to be

360 (optical path cantilever length = 45cm125um = 360) The relationship between laser

deflection and cantilever deflection is related by the mechanical amplification of the system

and thus the laser displacement is determined by

Laser Displacement = Cantilever Displacement Mechanical Advantage

Now applying a 10nm displacement to the cantilever and is it found that the afm laser will

deflect approximately 36 um on the pspd

36um (Laser Displacement) = 10nm (Cantilever Displacement) 360(Mechanical Amplification)

The XE-120 afm manufacturer PSIA states that the pspd itself is sensitive to laser

deflections of approximately 1nm 19 therefore the standard afm cantilever performs well

as a force sensor

Now we would like consider other cantilever lengths such as a 50 um cantilever and a 14

mm cantilever The 50um cantilever is potentially representative of an ion beam milled

cantilever and the 14 mm cantilever is representative of a tuning fork tine as cantilever

without modification By considering these two geometries one can compare how a

cantilever that has been ion beam milled into a tuning fork will perform versus how a tuning

fork tine unmodified will perform as a force sensor Using the same optical path (45 cm)

length for these cantilevers as before the mechanical advantages were found to be 900

(45cm50um=900) for the milled cantilever and 32 (45cm14mm=32) for the tuning fork

cantilever To make the comparison consistent we again apply a 10 nm cantilever

displacement to each cantilever and it is found that the expected laser deflection for a 50

um milled cantilever is 90um and the laser deflection for the 14mm tuning fork tine is 320

nm These three cantilever sensitivities are summarized in the table below

Table 4 Varying Cantilever Length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

Comparing the mechanical amplification of the standard non-contact afm cantilever of 125

um length with the ion milled cantilever of 50 um length a 250 increase in laser

deflection is observed with the shorter cantilever Comparing the standard afm cantilever

to the longer tuning fork tine a 91 decrease in laser deflection is observed for the longer

tuning fork tine This illustrates that a shorter cantilever is a more sensitive force sensor

In our experiment our cantilever is fabricated onto a QTF tine which subsequently

had to be mounted on the afm resulting in a variance in mounting height and thus a change

in optical path length Therefore it is of interest to examine the effect that an increased

optical path length will have on a cantilever sensitivity The cantilever and QTF was

mounted approximately 100um lower than the standard afm cantilever This results in a net

increase in the overall optical path length This is depicted in the figure 25 below a QTF

is mounted lower thus the optical path length is increased The table 5 below shows how

such an increase in optical path will change the total mechanical amplification for each

cantilever It is found that a change in 100 um of the optical path results in a 22 increase

in laser displacement

Figure 25 Illustrates how decreasing the mounting height of a cantilever in the z-axis results in an increase

in the optical path length (red dots) The left image depicts a standard afm cantilever mounted at the standard

height The right image depicts a tuning fork cantilever mounted at a lower z-height D with an increased

optical path length

Table 5 Varying Optical Path Length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

Comparing changes in cantilever length and optical path length it is found a change in

cantilever length will result in large changes in mechanical amplification while changes in

optical path length result in minimal changes in mechanical amplification Therefore since

the optical path length is rigid distance on a commercial afm the largest contributor to afm

sensitivity is cantilever length Based on these factors it was decided that an ideal

cantilever would have a length between 50-100um to maintain force sensitivity and

stiffness

Section 27 Lateral Spring Constant

The next parameter to investigate is cantilever width Cantilever width plays an important

role in afm measurements A typical afm non-contact mode cantilever width is between

22-37um (Nanosensors NCH-W) The cantilever width plays a key role in the lateral

spring constant of cantilever design Lateral spring constant designates the stiffness of a

cantilever to bending parallel to the scan direction Our experiment will involve a cantilever

that is being oscillated parallel to a sample surface therefore it will encounter a lateral

force as the probe interacts with the mesoscopic water layer and sample surface Our

interest is to characterize the normal force while generating an acoustic interaction with the

mesoscopic layer hence a cantilever that is rigid in the shear dimension is necessary Shear

motion is parallel to a sample surface The equation for the lateral spring constant has a

similar form as the normal spring constant except the thickness and width parameters are

exchanged The lateral spring constant is given as

Klat = tE4( w L )3 (EQ 25)

with Klat being the lateral spring constant t w L being cantilever thickness width and

length15 Notice that after having chosen a fixed range of cantilever lengths in the previous

section that the cantilever width has the largest overall impact on lateral spring constant

with a cubic power To tune the lateral spring constant for rigidity it was decided to

fabricate a cantilever on the higher side of the commercial cantilever width dimensions a

width of at least 35um would be fabricated In a later section once the cantilever thickness

is chosen we will estimate the lateral spring constant

Section 28 Laser Alignment and Interference

One potential design issue pertaining to cantilever width is the instance when the afm laser

spot size is larger than the cantilever is wide and spills over the edges of the cantilever and

onto the sample surface below15 This laser spot is then reflected off of the sample surface

and back onto the pspd creating laser interference As the z-piezoelectric is extended and

retracted a periodic laser signal of minimum and maximum intensities is observed16 A

typical approach curve as displayed in the figure 26 shows such laser interference The

standard method for remediating laser interference is re-aligning the laser on the cantilever

or repositioning the sample By using a custom cantilever care must be taken to fabricate

a device to minimize laser interference Our choice of fabricating a cantilever of a

minimum width of 35um should be sufficient at minimizing laser interference

Figure 26 A typical normal force approach curve that illustrates laser interference Sample substrate was

V1 grade Mica mounted perpendicular

Chapter 3 Cantilever and Tip Fabrication

We now have enough to fabricate a cantilever with a geometry that allows the

probing of the mesoscopic layer with both normal force sensitivity and rigidity in the shear

dimension for the generation of an acoustic signal The following sections will recap

equations describing cantilever stiffness and then the fabrication process for each

cantilever will be described As previously stated two separate designs were fabricated a

FIB milled cantilever into a tuning fork tine and a commercial afm cantilever was attached

to a tuning fork tine Both cantilevers had a Ga ion beam deposited carbon tip fabricated

on the cantilever to act as a geometrically repeatable probe

Section 311 Method 1-Focused Ion Beam Milled Cantilever Fabrication

To fabricate a cantilever onto a tuning fork tine a dual beam (Focused Ion

BeamScanning Electron Microscope) FIB-SEM FEI Helios 400s tool was used A beam

cantilever was ion milled into a tuning fork tine as shown in figure 31 The critical

dimensions of the beam cantilever were the length width thickness and tip height These

four parameters are used to tune the cantilever spring constants in the normal direction

lateral direction and torsional direction Only the torsional spring constant is dependent

upon the tip height and as such the tip height will be discussed in the tip fabrication section

Equations showing the relation between spring constants and cantilever dimensions

are shown in table 6 The milled cantilever beam dimensions are shown in table 7 A

cantilever length of approximately L = 76 um was chosen to enhance cantilever sensitivity

A cantilever width of approximately W = 35 um was used

To tune the cantilever spring constants for desired performance the only parameter

remaining was cantilever thickness t Ultimately a cantilever of 5 um thickness was

produced This resulted in a cantilever with a stiffer normal spring constant than a standard

tapping mode cantilever A tapping mode cantilever has a spring constant ranging from 20-

70 Nm while the fabricated cantilever has a normal spring constant of approximately 170

Nm This will allow the fabricated cantilever the ability to probe closer to the sample

surface while simultaneously delaying the jump-to-contact point closer to the surface The

lateral spring constant was found to be 10900 Nm and the torsional spring constant was

17959 Nm Both torsional and lateral spring constants are sufficiently large that

displacement in either dimension will be small compared to normal displacement Table 8

shows the material properties used for determining the spring constants and contains

material properties of the tuning fork (SiO2) and the ion beam deposited carbon used to

fabricate the tip Table 9 summarizes the estimated spring constants of the fabricated

cantilever

Table 6 Equations for Spring Constants

Normal Force Constant Kver Kver=wE4( t l )3

Lateral Force Constant Klat Klat=tE4( w l )3

Torsional Force Constant Ktor Ktor=w(G3)(t3l)1(H+t2)2

t=thickness w = width l= length H = tip height Source 15

Table 7 FIB Milled Cantilever Dimensions

Tuning fork Cantilever

Dimensions

Length (l) Width (w) Thickness (t) Tip Height

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

Table 8 Material Properties of QTF and IBD Carbon

Quartz Tuning Fork (SiO2) Youngrsquos Modulus

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

(Used as the probe tip material)

Shear Modulus G G = 70 GPa 22

Table 9 FIB Milled Cantilever Spring Constants

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

Section 312 Reflective Laser Pad

On the top side of the FIB milled cantilever an ion beam deposited platinum was fabricated

into a rectangle The role of this platinum pad was to increase the reflectivity of the

cantilever to increase laser reflection The deposition process entailed the use of the in-

chamber gas injection system The dimensions of this pad were approximately 3545um2

This platinum pad is shown in the figure 32-Image C Fabrication is shown in figure 31-

Section 313 Tip Base and Tip Fabrication

Afm tips have been fabricated using a FEI Helios 400s dual beam SEM-FIB with an

integrated carbon gas injection system (GIS) Tips with radius of ~100nm radius have been

fabricated using ion beam deposition Naphthalene C10H8 was the carbon gas source for

all carbon probe depositions (FEI Helios 400s User Manual) Before the carbon tip was

fabricated a rectangular platinum base was fabricated to serve as a structure onto which a

probe can be deposited A Gallium ion beam deposited platinum was deposited onto the

tuning fork using an in-situ platinum gas injection system constituting the rectangular tip-

base The platinum precursor gas was (methylcyclopentadienyl) trimethyl platinum

C9H16Pt (FEI Helios 400s User Manual) The rectangular base as seen in the figure 32-

Image E figure 33-image B below has a pyramid geometry The platinum base was

fabricated to be 8x8um2

The base was built by sequentially deposited platinum layers that were reduced in

width by 500 nm per layer deposition 13 layers of platinum were deposited Once the

width of the base was 15um gallium ion deposited carbon was applied The final tip was

deposited using spot mode ie the raster scanning was disabled Tip growth occurs rapidly

often taking less than 30sec The total height (base height + tip height) was ~5um in height

The final tip material was ion beam deposited carbon The current density for Pt deposition

is 2-6p Aum2 and the current density for Carbon deposition is 1-10p Aum223

Section 314 Electron and Gallium Ion Beam Deposited Carbon Tips

Electron beam carbon deposition onto tuning forks composed of the highly

insulating quartz material proved difficult due to electron charge build-up Charge buildup

is a problem when performing electron beam deposition on insulating materials Such

charge buildup results in a repulsive electric field developing on the tuning fork tine

causing electron beam induced tip growth to be skewed resulting in angled tips Proper

grounding of the tuning fork electrodes using silver paint (Ted Pella conductive silver

paint) allowed charge build-up to be reduced but not always eliminated

Due to this obstacle it was decided to produce Gallium ion deposited carbon tips

for experimental use Gallium ions are 127000 times more massive (Gallium 69723u

Electron 548510-4u) than electrons therefore charge build-up has a lesser deflective

effect on gallium ion beam path A gallium ion being opposite charge of an electron is

attracted to surfaces with built-up electron charge therefore both negligible beam

deflection results and the positive ions act to neutralize the negative charge build up Due

to the limited resolution of the ion beam tip radius of 100nm were reliable produced The

table 10 below shows the deposition parameters used for final carbon tip fabrication The

figures 31 32 33 the fabrication process for the FIB milled cantilever with a laser

platinum pad and (IBD) carbon tip

Table 10 Summary of Electron-Beam and Ion-beam Tip Deposition Parameters

Deposition Parameters Accelerating Voltage Beam Current Deposition time

EBD Carbon Tip 25 kv 25pA 65 sec

IBD Carbon Tip 30 kv 13pA 15 sec

Figure 31 A schematic illustration showing the tuning fork with a beam cantilever that was fabricated using

a gallium Focused Ion Beam (FIB) Image 1 shows the tuning fork with a corner of the tuning fork tine ion

beam milled away Red dots show the volume to be removed Image 2 Illustrates the tuning fork tine post

FIB milling Red dots illustrate another milled region that allows the cantilever the ability to flex in the

vertical direction Image 3 Illustrates a cantilever that flexes vertically Image 3B shows the cantilever range

of motion Image 4 illustrates the cantilever after an ion beam deposited platinum pad was deposited on the

top of the cantilever The purple color represents an ion beam deposited platinum pad used for laser reflection

Image 5 The cantilever is flipped over so that a carbon probe can be fabricated on the bottom of the cantilever

Image 6 and 6B illustrates the ion beam deposited probe (green color) onto the end of the beam cantilever

Figure 31-1 Schematic illustrates the FIB milling of the cantilever onto the tuning fork tine (A) The tuning

fork is aligned parallel to the ion beam (B) A beam cantilever is milled into the tuning fork tine (C) Shows

the resulting beam cantilever Ion beam (Accelerating voltage) currents ranged from (30kv) 65nA for rough

milling to (30kv) 1nA for fine polishing

Figure 31-2 Schematic illustrates the ion beam deposition of a reflective platinum pad on the top of the

beam cantilever (A) The tuning fork is rotated 180 degree so that the beam cantilever is aligned perpendicular

to the ion beam path with the top of the cantilever normal to the ion beam (B) The platinum gas injection

needle is inserted (C) A (Acceleration voltage 30KV) 5nA ion beam was used to deposit a reflective platinum

pad used for increasing laser reflectivity

Figure 31-3 Schematic illustrates the top side of the FIB milled beam cantilever before (A) ion beam

platinum deposition of the laser reflective pad and (B) after The platinum laser pad is shown in purple

Figure 31-4 Schematic illustrates the ion beam deposition of a platinum base and then a carbon tip The

tuning fork is removed from the SEM chamber and flipped over so that the bottom of the cantilever is now

normal to the ion beam (A) The tuning fork is aligned perpendicular to the ion beam (B) The

platinumcarbon gas injection needle is inserted A (Acceleration voltage 30KV) 25pA-1nA ion beam

currents were used to deposit the platinum base of the tip and subsequently the carbon tip

Figure 31-3 Schematic illustrates the bottom of the FIB milled beam cantilever (A) before and (B) after the

ion beam platinumcarbon deposition of the base and the tip (C) Is a close-up of the deposited tip

Figure 32 Schematic show the resulting FIB fabricated beam cantilever on tuning fork tine Images are

false colored to show material (A) (B) Bottom of the tuning fork (C) Gallium-Ion Beam Deposited (Ga-

IBD) platinum pad (purple) on top of cantilever for increased laser signal (D)-(E) Shows the first version of

a Ga-IBD deposited platinum base (purple) with carbon tip (green) Subsequent versions show improvement

in symmetry and height (F) Shows Ga-IBD carbon tip (green) with radius of 100 nm

Figure 33 Shows the refined iteration a platinum tip base and the final carbon deposited tip Image A

shows the FIB milled cantilever and a platinum pyramid base fabricated with Ga IBD deposition Image B

shows a higher magnification image of the platinum base (purple) and the carbon tip (green) Image C shows

the carbon tip A tip radius of 112 nm was achieved

Section 32 Method 2-Commercial AFM Cantilever Adhered to Tuning Fork Tine

This section will describe the fabrication of the second cantilever design To fabricate the

new design a commercial afm cantilever needed to be adhered to a tuning fork The figure

below shows a simplified description of how the tuning fork-afm cantilever combination

was fabricated Figure 34-Image A shows an unmodified afm cantilever and afm chip The

afm chip is a component attached to the afm cantilever to make macroscopic handling of

the cantilever possible otherwise the chip has no functional role in the afm data collection

Figure 34-Image B shows the process in which the afm cantilever is cut free from the chip

Figure 34-Image C shows the process in which the afm cantilever is welded onto the tuning

fork The tuning fork used was from manufacturer Abracon part number AB26TRQ-

32768KHz-T The AFM cantilever used was from manufacturer Nanosensors part

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

afm cantilever and afm

chip This is how it

arrives from the

manufacturer

Image B shows that

the afm cantilever is

cut free from the chip

Image C shows that

welded onto the tuning

fork tine

The actual fabrication process utilized a dual-beam FIB-SEM FEI Helios 400s microscope

The tool was equipped with a platinum and carbon gas injection system (GIS) that was

used for afm cantilever welding and tip growth The tool had an Omni-Probe system

integrated which is simply a platinum needle that has been etched down to a tip with a

micron diameter and has motor control in all three axes x-y-z allowing sub-micron feature

manipulation The Omni-probe system allows the user the ability to extract a sample from

a surface and reattach it at another location using electron or ion beam welding via gas

deposition After the commercial afm cantilever is adhered to the tuning fork tine a Ga

IBD carbon tip is fabricated on the cantilever using the same IBD carbon tip growth process

previously described

Figure 35 illustrates the removal process of a commercial afm cantilever from a cantilever

chip using the FIB-SEM FEI Helios 400s platinum GIS and an Omni-probe lift-out

procedure

Figure 36 illustrates the platinum gas injection welding of the afm cantilever onto a tuning

fork tine using the FIB-SEM FEI Helios 400s platinum GIS and an Omni-probe re-weld

procedure

Figure 37 shows SEM images of the fabricated device with platinum welds and Ga IBD

carbon tip labeled

Figure 35 Illustrates the extraction of the afm cantilever inside the vacuum chamber of a FIB-SEM FEI

Helios 400s (A) The Omni-Probe nanomanipulator is inserted above the afm chip and cantilever (B) The

nanomanipulator is lowered until approximately 05-1um above the afm cantilever surface (C) The platinum

gas injection needle is inserted (D) A platinum gas is injected into the chamber while the ion beam weld is

patterning This welds the nanomanipulator to the afm cantilever (E) Shows the resulting Ga ion platinum

weld (F) The Focused Ion Beam is used to cut the afm cantilever free from the afm chip The

nanomanipulator is raised 300um above the sample and retracted

Figure 36 Illustrates the welding of the afm cantilever onto a tuning fork tine inside the vacuum chamber

of a FIB-SEM FEI Helios 400s (A) The nanomanipulator is reinserted into the chamber above the tuning

fork tine (B) The nanomanipulator is lowered to a distance of approximately 05-1um above the tuning fork

tine (C) The platinum gas injection needle is inserted (D) A platinum gas is injected into the chamber while

the electron beam weld is patterning This welds the afm cantilever to the tuning fork tine Many platinum

welds were made to insure a secure attachment (E) Shows the resulting electron beam welds Electron beam

deposition was used due to irregular gas flow caused by the large afm cantilever geometry impeding platinum

gas flow Ion deposition attempts resulted in ion etching instead of deposition (F) Once welded the

nanomanipulator is FIB cut free then lifted and retracted

Figure 37 Above (A) Shows the actual tuning fork tines with an afm cantilever attached with electron beam

platinum welds (B) A higher magnification image of the afm cantilever Note the platinum weld at the center

of the cantilever was used to temporarily adhere to the nanomanipulator The afm tip is at the top near the

apex of the afm cantilever (C) Shows the platinum welds adhering the cantilever to the tuning fork tine (D)

A higher magnification image of the end of the cantilever (E) A top down image of the pyramid structure of

the afm with a carbon tip that was Ga ion beam deposited Tip deposition was the same as previously

described (F) A 52-degree tilted image of the IBD carbon tip

Table 11 Adhered Cantilever Dimensions

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

Table 12 Cantilever Material properties

Si3N4 AFM Cantilever Youngrsquos Modulus E E = 150 GPa 24

Table 13 Adhered Cantilever Spring Constants

735-219 Nm

Notice that only the vertical spring constant is calculated The lateral and torsional spring

constants were not presented due to the inability in identifying in literature the Youngrsquos

modulus of silicon nitride in alternate crystal orientations There was also found to be a

large variance in the published Youngrsquos modulus with values ranging from 130-150 GPa

Chapter 4 Formatting data and sampling techniques

Section 41 Force Distance Zero Line

Now that the cantilever fabrication process has been covered it is useful to review the

process of post-processing force-displacement data Three adjustments are required to have

meaningful data

First the force-displacement data captured from the experimentation process has

an arbitrary force offset applied to the acquired normal force data caused by a

vertical offset inherent to the laser positioning on the photodiode The offset is the

form of a DC voltage value resulting from the difference between pspd quads A-B

with non-zero values resulting from laser misalignment or laser drift as humidity

is increased This offset can be reduced by correcting the laser alignment before

experimentation but there always exists a small dc component that still must be

removed in post processing This process is called locating the zero force point

Simply stated this is removing the laser misalignment from force data

Second after the locating the zero-force point the zero distance point should be

found ie where is the surface The zero-distance point is defined as the

intersection of the zero force point with the force-displacement curve At this point

the cantilever is in an equilibrium contact with the sample surface and the observed

force on the cantilever is zero In other words this is finding z-displacement point

where there is tip and sample are in contact and there is zero force measured with

the cantilever

Third the force data needs to be scaled using the proper spring constant Most

commercial afms allow the user to select a commercial cantilever from a predefined

library however rarely do these values match the actual cantilever Since we

fabricated our own custom cantilever we need to rescale the force appropriately

Section 42 Normal Dimension Zero Force Point

Following the work of Cappella et al16 it is established that the zero point (Fn=0) of normal

force component is defined as being a distance far from the sample surface where there is

essentially no observable normal force present on the tip In our experiments we will

define the zero point distance as approximately 50 nm from the surface (Fn=0 Z=50nm)

The justification behind this can be found using the simple sphere-plane (SP) model of the

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

with H being the Hamaker constant R is the tip radius and Z is the z-axis height from the

sample surface Letting an estimated Hamaker constant be H = 510-18 J R = 100 nm using

a tip to sample height of Z = 50 nm the observed force from the afm tip and surface

interaction is FN = -22 pN an attractive force To compare this to thermal noise at ambient

conditions we can reference Cappella16 the thermal fluctuations in the normal component

of force Ft

Ft le 30pN

These thermal fluctuations are larger than the observed surface-tip interaction forces at 50

nm in height It is plausible to assign the normal force zero point at a distance gt= 50nm

from the surface The zero-force point is marked as a horizontal line A on the plot shown

figure 41 below It intersects points on the force curve where the cantilever experiences

zero force

Section 43 Z-height Zero Point

The zero point (Z0=0) of the z axis is defined as the point where the force curve

intersects with the horizontal line A For improved clarity this is illustrated in the figure

41 above as point B At point B the tip and sample are in direct contact and the observed

force is zero

Figure 41 A schematic

representation of the approach

and retraction curve The

point A represents the zero

point of the normal force

component The point B

represents the zero point of the

z-axis distance component

Notice that the force at point B

is zero The definition of the

surface is accepted to be the

intersection of measured force

curve and the zero normal

force line [1] The point C is

the snap-to-contact point

Image from 27]

Section 44 Normal Force Scaling

To scale the experimentally obtained data one simply multiplies the observed force

by the theoretical spring constant and divides by the experimentally observed spring

constant The experimentally observed spring constant is the slope of the force curve in the

contact region The theoretical spring constant is obtained from SEM cantilever

measurements Corrected force has the following form

Fc = Fm Ktheory Kmeasure (EQ 42)

with Fc being the corrected force Fm being the measured force Ktheory being the estimated

cantilever spring constant from SEM measurement and Kmeasure is the slope of the force

curve in the contact region

Section 45 Cantilever Deflection Amplitude

In shear force experiments it is useful to have an estimate of tuning fork amplitude

in terms of actual displacement distances Traditionally this is measured using

interferometric techniques In this thesis we attempt to directly measure the tuning fork

amplitude displacement by mechanically contacting a vertically oscillating tuning fork tine

with an afm tip while simultaneously monitoring the laser deflection voltage We are able

to convert the laser deflection voltage into a tuning fork displacement distance by obtaining

a scalar that relates laser deflection voltage to tuning fork deflection distance The

parameter relating laser deflection voltage to cantilever deflection is known as the

deflection sensitivity In section 46 we will briefly cover deflection sensitivity in

anticipation of performing such experiment in chapter 5 to estimate tuning fork deflection

distances as a function of tuning fork drive voltage

Section 46 Deflection Sensitivity

To obtain the deflection sensitivity parameter a laser deflection-displacement

curve is needed A laser deflection curve is obtained in the same manner as a force curve

with the difference being that the afm records a voltage that is proportional to the position

of the laser beam spot as it is deflected across the position sensitive photodiode as the z-

piezoelectric is extended toward the sample surface The parameter known as the deflection

sensitivity β has units of nmV Distance here is cantilever displacement and volts are the

output from the pspd The deflection sensitivity values are not readily available to afm

users therefore a method to obtain them is necessary The deflection sensitivity parameter

can be obtained by performing a laser-displacement curve (A force curve while observing

the laser voltage instead of force) and recording the slope in the contact region In this

contact region the deflection of the cantilever is directly proportional to the z-piezoelectric

displacement The slope of the laser-displacement curve in the contact region has units of

Vnm Taking the inverse of this one obtains the deflection sensitivity parameter β

Typical values of β range from 10-100 nmv 2728

Section 47 AFM Scanner Calibration

The process to obtain a meaningful approach and retraction curve requires a

properly calibrated tool For reference to the full scanner calibration procedure please see

afm manual from Park (Park Systems Corporation XE-120 High Accuracy Biological

Sample SPM User Manual Version 184 (2013) p 127 )19 A properly calibrated tool

requires that the z-axis be calibrated with respect to a known distance To calibrate this

dimension a cleaved piece of silicon wafer (2cm2) was masked with a piece of tape and

partially sputtered with iridium to a known thickness (135nm) This value was verified

by performing a scanning transmission electron microscope (STEM) measurement using

an FEI Helios 400s figure 44 For x-y calibration a checkerboard patterned calibration

stub (MicroMesh TGX series 1 10um pitch) was used

The XE120 AFM operates the x y z scanning piezo electrics in low voltage mode

and high voltage mode The XE-120 afm uses a 16bit digital-to-analog converter (DAC)

which allows 216 65536 addresses In high voltage mode the z piezoelectric has a

displacement of 12um and in low voltage mode it has a displacement of 17um By

operating in two separate modes the afm is able to achieve resolutions of 183Å and 026Å

for high voltage and low voltages respectively This z-resolution is easily found with

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

ie 12um 65536 voltage addresses = 183Å = Z high voltage resolution

17um 65536 voltage addresses = 026Å = Z low voltage resolution

In high voltage mode the x-y piezoelectrics have a displacement of 100um and in low

voltage mode it has a displacement of 10um The effective resolutions for x-y

piezoelectrics for high voltage and (low voltages) modes are 152Å and (152Å)

respectively (XE-120 manual pg 71)19 The XE-120 manual recommends that the afm be

calibrated in the size scales that will be used within the experiment therefore the z-

dimension should be calibrated with atomic dimensions if possible The x-y scanner will

not be used in our experiments but it was calibrated within the micron domain for

consistency The calibration procedure was to obtain a topography image while

maintaining proper tracking For the z-axis calibration since the measurement was sub-

nanometer in scale a fresh tip was used If an old tip is used a dull tip will not be able to

properly sample small cracks and hole-like features resulting in the substrate appearing to

be smoother than actuality Images were obtained with a tapping mode cantilever

(Nanosensors NCH-W) with a spring constant of approximately 40Nm

Figure 42 High Voltage Z-

Calibration Profile

Measured Profile was found

to be 2173nm Accurate

value was expected as 1350

nm +-05nm Notice the

large perturbation (~75nm

height) at the interface

between the iridium coated

sample region (right) and the

bare silicon region (left)

Using the STEM procedure

this was found to be a glue

artifact from the masking

procedure

Figure 43 Low Voltage Z-Calibration Profile Measured Profile was found to be 140nm Accurate value

was expected as 1350 nm Notice the large perturbation at the interface between the iridium coated sample

region (right) and the bare silicon region (left) Using the STEM procedure this was found to be an artifact

from the masking procedure

It is interesting to note the large (height) perturbation at the interface between the iridium

coated sample region (right) and the bare silicon region (left) in the figures 42 amp 43 above

The z-axis calibration stub was fabricated by applying scotch tape as a mask to half of the

sample and then coating the sample with iridium In theory removing the mask afterwards

would result in a step height interface suitable for atomic calibration Scanning

Transmission Electron Microscopy (STEM) procedure was applied to verify the iridium

coating thickness In the STEM images shown below the large perturbation appears bright

indicating that the material is a low atomic weight material Notice that the perturbation is

embedded beneath the iridium coating This supports the conclusion that the perturbation

is lsquogluersquo from the masking procedure By using a large scan area where the glue regions

are disregarded a step height between the clean silicon and iridium coating can be obtained

to calibrate the z-axis of the afm

Figure 44 The STEM image shows the interface between iridium and bare silicon z-axis calibration

standard In the center of the image is a large low-z material lsquogluersquo embedded under the iridium coating This

suggests that it is from the masking procedure (1) Protective Electron Beam Deposited Platinum (2)

Protective Electron Beam Deposited Carbon (3) Bare silicon substrate (4) Low-z material (5) Iridium coating

is the thin black line

Figure 45 A high

resolution STEM

image of the low-z

lsquogluersquo perturbation It

measures 74nm in

height AFM

measurements

indicate a 75-100nm

height of the interface

perturbation

Figure 46 High Voltage X-Y Calibration Profile Measured Profiles were x = 39919um and y = 4002um

Expected profile was (x-y) 4000x4000um Feature spacing was 10um

Section 48 Measurements with a Lock-In Amplifier

To make a measurement of an oscillating tuning fork amplitude or an oscillating acoustic

signal one needs to make use of a lock-in amplifier For a thorough introduction to lock-

in amplifier theory it is recommended that one reads literature covering the subject The

SR850 instruction manual provides key insights into lock-in theory29 The university of

Pittsburgh has compiled a great introductory tutorial covering the subject30 Following

DeVore et al from the University of Pittsburgh in deriving the output of a lock-in

amplifier the expected output of a lock-in amplifier before a low pass filter has the form

Vmx = 12G cos( 2π(Fr-Fs)t + φ)+ cos( 2π(Fr+Fs)t + φ) (EQ 43)

Vmy = 12G sin( 2π(Fr-Fs)t + φ)+ sin( 2π(Fr+Fs)t + φ) (EQ 44)

where Vmx (Vmy) is the voltage in the x-quadrant (y-quadrant) after the mixermultiplier

G is the gain of the amplifier Fr is the reference signal Fs is the experiment signal φ is

the phase delay Notice the ideal case of Fr=Fs the experiment signal exactly matches the

reference signal In this case Fr-Fs is zero and Fr+Fs=2Fr This gives an output of

Vmx = 12G cos(φ)+ cos( 2π(2Fr)t + φ) (EQ 45)

Vmy = 12G sin(φ)+sin( 2π(2Fr)t + φ) (EQ 46)

where the first term in each expression is a DC component and the second term in each

expression is an AC component that rapidly oscillates To obtain only the DC component

a low pass filter is used such that the higher 2Fr component is filtered On a lock-in

amplifier to control filtering one has the ability to designate a time constant that is

proportional a cut-off frequency This relation has the form

Τ = 1(2 π Fc) (EQ 47)

where T is the time constant Fc is the cut off frequency at -3db In our experiments we

use a reference frequency around 32768hz (+- 300hz depending on geometry and mass

of attached tip) the 2f component is approximately 65kHz To filter the 2f component

would require a cutoff frequency below this value Letting a cutoff frequency be 60kHz

one would choose a time constant of T = 265 us Any shorter time constants would result

in passing for the 2f component Lowering the cutoff frequency has the effect of increasing

the time constant For long time constants on the order of 500ms the cutoff frequency is

031Hz Such long time constants are able to remove noisy signals around our frequency

of interest but at the cost of adding a long time delay to our signal Further such long time

constants result in a slow response to rapidly changing signals In the experiments

described in sections 55 amp 56 we use a short time constant of 3ms to filter out the 2f

component of the tuning fork signal

Section 49 Lock-in Amplifier Sampling Rate

The z resolution of signals tracked with a lock-in amplifier are directly proportional to

sample rate and afm approach speed Our afm approach velocity is limited to a minimum

approach speed of 10 nmsec To obtain a z resolution of 05nm would require a sample

rate of 20 samples per second A lock-in amplifier needs 5X the time constant T to obtain

a single data point This 5X is known as the settling time or the averaging time Therefore

we can estimate the maximum time constant based on the above sampling rate and

approach velocity A single data point should be sampled every 120th of a second but the

sampling procedure takes 5 T seconds Therefore the maximum time constant is ⅕ of

the sampling rate 1(Sample Rate5) seconds or 10ms

Section 410 AFM Noise Level

The noise level of the XE-120 afm was obtained by analyzing an afm force approach curve

in the contact region In the contact region the slope of the force curve follows Hookes

law and is constant By performing a linear fit in this region and removing the slope from

the experimental data we are able to project the noise into either the normal force axis or

in the z-axis Plots below show the resulting noise signals in each axis The standard

deviation of the noise is used as measure of the noise level It was found that there exist a

022nm noise level in the z-axis and a 11nN noise in the normal force Please note this is

a considerable noise signal while small compared to our experimental findings it could

possibly be caused be a result of an environmental noise an unbalanced air table or both

Figure 47 Shows z-axis afm noise Shows the resulting contact normal force data projected onto z=0 The

standard deviation of the z-axis is 022nm

Figure 48 Shows afm normal force noise Shows the resulting contact normal force data projected onto

Fn=0 The standard deviation of the normal force-axis is 11nN

Chapter 5 Experimental Procedures

The purpose of chapter 5 is to experimentally verify four key properties related to the

investigation of surface forces The four experiments that were performed are

Section 51 List of Experiments

1 Experiment 1 An experiment to measure the tuning fork lateral displacement using

an afm laser-displacement curve The goal is to estimate the tuning fork

displacement as a function of driving voltage

2 Experiment 2 AFM force curves are obtained using a non-contact afm cantilever

with the intent of directly observing the snap-to-contact vertical position By

observing the snap-to-contact position it is possible to obtain an estimate for the

Hamaker constant for a Carbon tip and a Mica surface

3 Experiment 3 EDX analysis is performed on the electron and ion beam deposited

carbon tip material Material is deposited as a thin film and then characterized with

edx to determine the relative elemental concentrations

4 Experiment 4 The FIB milled cantilever is driven to oscillate parallel to the sample

surface and is interacted with a sample surface to generate normal force tuning fork

amplitude and acoustic signals

5 Experiment 5 A laterally oscillating tuning fork with a commercial afm cantilever

attached is interacted with a sample surface to generate and observe normal force

tuning fork amplitude lateral displacement and acoustic signal

Section 52 Experiment 1 Directly Measuring Tuning Fork Oscillation Amplitude

Previous shear force experiments by Karrai and Gregor used interferometry to

observe tuning fork displacement distance We will attempt to directly measure tuning fork

displacement distance using a direct contact afm method We will compare measure tuning

fork displacements with the interferometric results to assess the validity of such approach

To measure the tuning fork displacement distance using an afm a tuning fork was mounted

vertically with the oscillatorrsquos motion constrained up and down in the z-axis as seen in the

figure 51 below A non-contact NanoSensor (NCH-W) afm cantilever with spring constant

of ~40 nNnm was then contacted with the tuning fork while the tuning fork was driven at

resonance The tuning fork tines were mounted perpendicular to the afm tip and were

electrically driven to oscillate up and down in a vertical motion causing the afm tip to

deflect vertically By obtaining a deflection sensitivity value it is possible to convert the

observed afm laser deflection (voltage (v)) values into tuning fork deflection (distance

(nm)) values The tuning fork was mounted using cyanoacrylate onto a aluminum cylinder

of height approximately 8-10 mm A small cylinder was chosen to maintain sufficient afm

tip clearance between the afm head and the sample mount

Figure 51 Image A Illustrates an afm cantilever in direct contact with a stationary tuning fork Image B

illustrates an afm cantilever in direct contact with a tuning fork electrically driven to oscillate in the vertical

direction This vertical tuning fork deflection results in a laser deflection proportional to tuning fork

deflection amplitude

The tuning fork (TF) current (I1) and driving voltage (V1) were observed with a SR850

digital lock-in The tuning fork oscillation amplitude was directly observed by operating

the afm in contact mode without feedback There is concern that operating with feedback

would result in the feedback loop constantly adjusting to maintain the force set point

negating tuning fork deflection amplitude measurements This concern can be relaxed since

the tuning fork operates at 32 KHz a much faster rate than the feedback loop The feedback

loop would in essence observe the cantilever as being stationary Operating in direct

contact without feedback could result in a linear drift in the observed laser voltage as a

sloping DC voltage This slope could be measured and subtracted from the data

Figure 52 Schematic illustrates experiment setup for measuring tuning fork displacement A digital lock-

in (SR850) is used to electrically excite a tuning fork The tuning fork is oscillated vertically causing vertical

displacement of the afm cantilever The afm cantilever displacements are observed by monitoring the laser

deflection with the pspd of the XE-120 afm The (PSPD Quads A-B) vertical laser deflections of the pspd

voltage (AC) component is passed through a signal access module (SAM) with a standard BNC to a Tektronix

TCD 2024B oscilloscope for observation

During the experiment the afm cantilever is directly mechanically coupled to the tuning

fork and oscillates at the same frequency as the tuning fork By isolating this frequency on

an oscilloscope it is possible to measure the laser (AC) deflection voltage Vpp However

it is not a voltage value that we are interested in but the actual deflection height of the

tuning fork therefore the deflection sensitivity (β) will be needed

To pass and record the laser voltages on the oscilloscope the laser deflection

voltages were transferred from the XE-120 afm to the oscilloscope using an intermediary

breakout box known as the XE Signal Access Module (SAM) The SAM allows users the

ability to access control voltages supplied to and from the hardware of the afm It allows

the user to act as a ldquoman in the middlerdquo and gives them the ability to copy critical drive

signals to external hardware for custom experiments The SAM was used to pass the

vertical deflection component from the PSPD to a Tektronix TCD 2024B oscilloscope

where it was recorded The oscilloscope was triggered externally by the lock-in amplifier

driving the tuning fork For clarity the vertical deflection components were the voltage

difference between quadrants A and quadrant B of the pspd As previously mentioned A

= a+b and B = c+d where a b are the upper two quadrants on the pspd and c d are the

lower two quadrants on the pspd See Figure 11 for clarity of pspd functionality Therefore

subtracting the upper quadrants from the lower quadrants will give a voltage difference

proportional to vertical laser deflection The range of output voltages from the SAM were

+-10 V

The sensitivity parameter is obtained by performing a force distance curve on a

non-deformable hard sample In our experiment the laser deflection voltage is recorded

instead of force See figure 53 for a laser deflection voltage vs distance plot (The recorded

voltage here is the laser deflection voltage induced by the cantilever deflection and should

not be confused with the driving voltage of the tuning fork which is recorded with the

SR850 lock-in amplifier) This will result in a laser deflection voltage versus displacement

curve (LDV-D) that would have slope units of Vnm Deflection sensitivity is the inverse

of the slope of the LDVD-D curve with units of nmV The single crystalline quartz tuning

fork was used as a hard sample surface to obtain β It should be noted here that sensitivity

values depend upon cantilever mounting height afm cantilever length and laser spot

positioning on the cantilever Therefore for accurate scaling of cantilever deflection

voltages it is recommended that one should obtain a deflection sensitivity value after every

direct displacement experiment Tip damage can occur with direct contact so deflection

sensitivity parameters should be obtained after the experiment has been performed

A value of β = 2248 nmV was observed Typical sensitivity values range from 10-100

nmV31 In the figure below a laser-displacement curve is plotted to find the deflection

sensitivity parameter notice that only the contact region is plotted this is so that a linear

fit can be used to obtain an inverse slope value for the deflection sensitivity parameter

Figure 53 A laser

deflection voltage to

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

Multiplying the deflection sensitivity β by the laser deflection voltages Vpp that were

observed with the oscilloscope a TF displacement distance is obtained The TF

displacement distance is plotted against the recorded tuning fork drive voltages and drive

currents in the figure 54 below The plot to the left in figure 54 is the TF displacement

distance versus measured TF current and the plot to the right is the TF displacement

distance versus TF drive voltage At the lowest drive voltages 4mV a TF displacement of

approximately 1nm was observed At higher drive voltages 50mV a TF displacement of

approximately 54nm was observed A linear relationship for (TF-displacement versus

drive-voltage) and (TF-displacement versus TF-current) is observed It is found that for

every increase of 10mV drive voltage there is an increased tuning fork displacement of

approximately 1nm or simply 100 pmmV Previous work by Grober et al have directly

measured physical tuning fork amplitude displacements of 596+-01 pmmV from

interferometric measurements32 Our experiment used a different tuning fork than Grober

et al but the comparable results seem to suggest that direct mechanical observation using

an afm cantilever yields a good estimate of the tuning fork displacement

Figure 54 Demonstrates relation between tuning fork displacements and measured electrical response

(Left) Shows the relation between TF displacement and measured TF current (Right) shows the relation

between TF displacement and TF drive voltage Notice the linear relation between observed tuning fork

displacement and observed electrical characteristics

It should be noted that this TF displacement distance is a reduced value By the nature of

the experiment the afm cantilever applies a small restoring force to the tuning fork

reducing the total amplitude to some reduced amplitude The amount of force applied can

be estimated by applying Hookes law F=-kx with k being the afm cantilever spring

constant and x being the total observed displacement With K~40nNnm and x ranging

from 1-20nm the applied force ranges 40-800nN For future cantilever deflection

measurements it is recommended to use low K cantilevers such that only small restoring

force is applied and thus limited tuning fork damping occurs Contact mode cantilevers

have spring constants well below 1nNnm such as K = 005nNnm If used a maximum

force of approximately 1nN could be expected while driving a tuning fork amplitudes of

20nm However even with stiff cantilevers such as the one used in our experiment as long

as the deflection distances are small so too will be the resulting dampening force

Section 53 Experiment 2 AFM Normal force curves to obtain Hamakerrsquos Constant

This section covers the experimental process for obtaining a normal force curve A force

model will be used to obtain an estimated Hamaker constant for an IBD carbon tip with

radius of 100nm with a mica surface at high humidity (gt40)

The force model used follows the work of Soma Das et al17 The procedure proposed by

Dasrsquos group utilizes a force curve interaction to directly measure the Hamaker constant by

deriving a relation between the snap-to-contact point of an afm force curve the tip radius

the cantilever spring constant and the Hamaker constant They make use of a simplified

model of the afm force interaction using the sphere-plane (SP) approximation to represent

the afm tip and sample substrate The relation obtained by Das is

A = (2427)(KcRt) hj3 (EQ 51)

where A is the Hamaker constant Kc is the normal spring constant Rt is the tip radius and

hj is the experimentally measured jump to contact position This function is valid in the

domain of zgta0 with a0 being an atomic distance generally defined as a0 = 015nm By

operating at heights larger than 015 nm the system is well approximated by Van der Waals

forces A note of caution at high humidities there is a mesoscopic water layer present and

such model is invalid within contact with this region therefore the Hamaker constant will

only be obtained for regions above the mesoscopic water interaction

Section 531 Experimental Force Curves

A series of force curves were obtained using a regular afm cantilever with a spring constant

of Kc = ~40 nNnm A IBD carbon tip with a radius of 100 nm was fabricated onto the afm

tip using the ion deposition procedure previously described A maximum force set point of

100 nN was defined A 10mm v1 grade mica surface was used (Ted Pella Inc Highest

Grade V1 AFM Mica Discs 10mm) Humidity was 45-47 The XE-120 afm was

operated in force-displacement mode

The force curve data presented below uses the method previously described in section 42

to define the zero force point and then uses the intersection of this force height with the z-

displacement to find the zero-distance point as well A well-defined snap-to-contact point

at Z0 = 25 nm and Fn = -70 nN is observed Using the jump-to-contact point in conjunction

with the (EQ 51) it is possible to estimate the Hamaker constant This is illustrated in the

table 14 below

Figure 55 This figure shows the observed normal force with the zero force and zero distance points Points

of interest labeled with solid black horizontal and dotted vertical lines The snap-to-contact point is labeled

and found to be 257 nm with an attractive force of 70 nN The approach curve is then parsed into a region

before surfacemesoscopic interaction (blue line) and a region after mesoscopic layer and surface interaction

(red line)

Section 532 Estimated Hamaker Constant

Table 14 Hamaker Constants for IBD Carbon Tip and Mica Surface

Hamaker using Snap-to-contact Model

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Parameters used in EQ 51 hj = 257-271nm Rt =100nm Kc~40Nm

These measured values for the Hamaker constant are reasonable Typical Hamaker

constants are generally of magnitudes (10-21 to 10-18) J Examining the obtained force

curves shows a perturbation at approximately the 5-7nm position which is caused by the

probe-mesoscopic layer interaction to improve the accuracy of this Hamaker constant

future experiments should make use of a contact mode cantilever (low spring constant

cantilever) to reduce the interaction duration with the mesoscopic water layer Lower

humidity conditions would reduce contributions of the water layer to the observed normal

Section 533 Probe Reliability

To investigate the probe reliability a probe was used to acquire 10 normal force curves

without a shear oscillation applied Normal force curves were acquired to obtain the

Hamaker constant between a mica surface and a IBD carbon tip Before and after SEM

images show that tip wear was minimal Figure 55-1 When a shear force was applied tip

wear was observed Figure 55-2

Figure 55-1 Left SEM image is before and right is after acquiring 10 normal force curves without a shear

force applied A force set point of 100nN was used 10 force curves were acquired to obtain the Hamaker

constant between a mica surface and IBD carbon tip

Figure 55-2 Image shows an example of probe wear after many (10rsquos to 100s) normal force approach curves

while a shear oscillation is applied The final probe is approximately 415nm in radius

Section 534 STEM Analysis of IBD carbon

A STEM analysis of IBD carbon was performed See literature for STEM preparation

procedure The figure below shows the resulting bright-field image of the grain structure

from the ion beam deposited carbon Notice that the IBD carbon structure shows

nanoparticles and an amorphous structure is present The nanoparticles are estimated to be

between 1-5 nm in diameter without long range structure It is not known if these particles

are diamond like carbon

Figure 55-3 A bright-field STEM image showing the nano-structure of IBD carbon

Section 54 Experiment 3 EDX characterization of EBD and IBD carbon materials

Section 541 EDX Characterization of Electron and Ion deposited Carbon

An energy dispersive x-ray spectra (EDX) was acquired to characterize the ion

implantation that occurs while performing a Ga ion carbon tip deposition A series of

carbon thin films were deposited onto a freshly cleaved silicon wafer using both electron

beam deposition (EBD) and ion beam deposition (IBD) using an FEI Helios 400s SEM-

FIB dual-beam with an integrated naphthalene gas injection system and a Bruker

instrument spectrometer with 133 eV resolution

An EDX signal is composed of two main components the bremsstrahlung x-rays which

are considered background noise and the characteristic x-ray peaks which are used for

elemental identification Bremsstrahlung noise is the generated when electrons and x-rays

scatter as they interact with a substrate lattice A Bremsstrahlung signal is an x-ray signal

that is generally lower in x-ray counts and covers an energy range that broadly spans the

total sampled energy spectrum Characteristic x-rays are the result of direct electron

transitions that occur when an electron fills a shell vacancy left after the incident electron

beam scatters a lattice electron X-rays that make it directly to the detector without energy

loss contain information that is unique to the scattering atom thus they are ideal for atomic

elemental identification

Electron beam accelerating voltage and beam current play a critical role in the EDX

characterization of unknown specimens Electron beam accelerating voltage controls the

maximum excitation energy and beam penetration depth Both the standard 15 kv and also

a lower 5 kv acceleration voltage were used At 5kv many atomic elements have unique

characteristic x-rays available for identification A 15kv acceleration voltage is able to

probe higher electron energies and is used to verify elemental identification by exciting

higher energy transitions

Section 542 Electron Beam Acceleration of 5k Volts

Choosing a 5 kv acceleration voltage was an attempt to limit the sample penetration depth

to only the carbon film and to observe the lower K-alpha energies of carbon silicon and L-

alpha of gallium A 5 kv acceleration voltage produced a spectra with low x-ray counts and

subsequently a high degree of uncertainty in elemental identification

The 15 kv beam provided sufficient x-ray counts for reliable elemental identification and

also a direct excitation of both L-alpha and K-alpha energies in Gallium

Table 15 contains the beam conditions used for thin film depositions of electron beam and

ion beam deposited carbon The last row of table 15 shows the beam conditions

(acceleration voltage beam current and sampling time) used for edx characterization

Table 16 lists the excitation energies for the elements identified from EDX Table 17

contains a summary of atomic percentages from each thin film edx sampling

Table 15 Summary of Electron-Beam and Ion-beam thin film Deposition Parameters

Deposition Parameters Accelerating

Voltage

Beam Current Deposition time

EBD Thin Film Carbon 20 kv 11nA 300 sec

IBD Thin Film Carbon 30 kv 30pA 300sec

EDX Parameters

(Spot sampling)

5 kv and 15 kV 14nA EDX sample time 120

Table 16 Energy Table for EDX analysis

Carbon (C) Silicon (Si) Gallium (Ga) Aluminum (Al) Tin (Sn)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

Table 17 Summary of EDX characterization

Material Atomic Carbon

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

Section 544 EDX Results

It was found that electron beam deposited carbon thin-films were 99 carbon if

impurities are ignored and 976 pure if Al and Sn impurities are considered It was found

that Gallium ion beam deposited carbon contained 25 gallium and 75 carbon with no

impurities detected All results disregard the background Silicon signal

Comparing the EBD and the IBD edx results one can conclude that the naphthalene

gas was of high purity gt976 This conclusion is obtained by noting that the IBD signal

contained no impurities while the EBD signal contained trace impurities A possible source

of impurities was the silicon substrate The penetration depth of electrons at 15 kV is 18um

in carbon34 given by the Kanaya-Okayama range It should be noted that the EBD carbon

film thickness was less than the Kanaya-Okayama range and since a strong silicone peak

was observed it can be concluded that sampling of the silicon substrate occurred Since

IBD carbon deposited faster12 a thicker carbon layer was formed for similar deposition

times therefore it is plausible that there was less sampling of the silicon substrate This is

supported by examining the EDX spectra of both EBD and IBD It is clear that a larger

silicon peak is observed for the EBD edx spectra

Since it was observed that there were no trace elements found in the IBD edx analysis this

supports the idea that the trace elements were contaminants of the silicon substrate and not

the carbon precursor gas Inspecting the elements in the energy table for edx analysis it

can be observed that all elements are well separated in energy excitation levels thus it is

improbable that the gallium peak is obscuring the trace elements Again disregarding the

silicon peak as background one finds a carbon concentration of 75 and a gallium

concentration of 25 This corresponds well with observed values in literature Lemoine

et al12 which observed values of gallium implantation in IBD carbon of 20-25

Following the work of Lemoine et al they observed the material properties of IBD carbons

tips to have a Youngrsquos modulus between E=120-130 GPa An intrinsic hardness was

observed to be approximately H=9-10 GPa Comparing the Hardness values of silicon (9-

124 GPa)35 and mica (512 GPa)36 with that of IBD carbon (9-10 GPa) it is found that

IBD carbon is softer than crystalline silicon but harder than mica

Figure 56 Shows the EDX spectra for IBD carbon (Top) and EDB carbon (Bottom)

Section 55 Experiment 4 Surface interaction using FIB milled Cantilever

Section 551 Experimental design

The objective of this experiment is to observe simultaneously the normal force and

shear force while obtaining approach and retraction curves using a custom designed

cantilever and probe The tuning fork ultrasound signal was also observed using an acoustic

emission transducer This experiment will use a cantilever that has been FIB milled into a

tuning fork tine Results will discuss the observed findings and experimental limitations

Section 552 Measuring Normal Force

To observe the normal force a rectangular cantilever was focused ion beam milled

into the tine of the 32768khz tuning fork and a carbon tip was ion beam deposited onto the

cantilever as previously described The milling process removed mass from the tuning fork

increasing the resonant frequency to 33109khz The tuning fork and probe were then

mounted (adhered with cyanoacrylate) on the AFM such that the tuning fork tines would

oscillate parallel to the sample surface see figure 57 below The XE-120 afm laser was

aligned top of the cantilever such that interactions with the normal forces would cause a

direct vertical laser deflection on the A-B quads of the position sensitive photodiode The

XE-120 electronics will record the normal force data as the force curve is obtained The

vertical spring constant of the FIB milled cantilever was found to be approximately 170

Nm For calculation of the spring constant see section 311 table 6

Figure 57 Illustrates the tuning fork oscillating motion relative to the sample surface (Image 1) This

oscillation motion is constrained to be parallel to the sample surface in the x-axis Image 2 illustrates a

cantilever before contact with the sample surface Sides A BC are labeled for dimensional clarity Image 3

illustrates a cantilever interacting with a sample surface and bending as force is applied Notice the laser

deflection as the cantilever bends

Section 553 Measuring the Shear Force

To observe the shear force the tuning fork was driven electrically with a digital

lock-in amplifier (Stanford Research SR-850) The tuning fork was driven at 50mV

estimated 5nm amplitude While being driven the piezo-electrically generated current

from the tuning fork was observed using the same digital lock-in amplifier The drive signal

was generated and phase locked monitored using the internal frequency source of the lock-

in amplifier Sensitivity of the lock-in amplifier was set to 1uA A time constant of 3ms

was used The SR-850 digital lock-in then output the magnitude (R) of the measured tuning

fork current as a DC voltage through channel 1 output This DC voltage is then passed to

the analog-to-digital-converter (ADC1) on the XE-120 afm using a standard bnc Channel

1 has an output range of +-10v with a maximum 10mA output The ADC1 accepts an

input of +-10v with a 50kΩ impedance

Section 554 Measuring the Acoustic Response

The commercially available acoustic emission transducers AET R15 (Physical

Acoustic Group R15a) was used to record the acoustics generated from the tuning fork In

the following experiment a different acoustic sensor is used specifically the SE40-Q

(Score Dunegan SE40-Q) This was an attempt to improve the acquired acoustic signal In

the current experiment the R15a acted as a bottom sensor ie it was mounted beneath the

substrate as seen in the figure 58 below The acoustic signal was monitored using a second

digital lock-in amplifier (SR-850) The SR-850 used an external trigger from the first lock-

in amplifier The measurement was a phase locked current measurement with a sensitivity

of 20nA A time constant of 3ms was used The acoustic sensorrsquos current was sampled

similarly to the tuning fork signal it was passed from the channel 1 output of the lock-in

amplifier into the ADC2 on the XE-120 afm ADC2 has the same power considerations

as ADC1

Section 555 Substrate Preparation

The substrate that was used was a 1cm2 silicon wafer cleaned with ipa A native oxide layer

was present on the silicon substrate The edge of the sample (within 20um of edge) was

used for force curve collection This was an attempt to avoid unwanted contact of the tuning

fork and the substrate It is only desired that the probe tip and substrate have interaction

The sample was adhered to the acoustic emission sensor using vacuum grease (Dow

Corning High Vacuum Grease)

The afm is enclosed inside of an acrylic enclosure with an air feed through line to control

atmospheric humidity Humidity was increased from ambient conditions (10-45

humidity) to approximately 60 by pumping air at 12-15 psi through a bubbler in a beaker

of deionized water Humidity was maintained between 55-65

Figure 58 Experiment Schematic illustrating the normal force shear force and acoustic interaction

measurements The tuning forkprobe was mounted on the XE-120 afm in place of a regular afm cantilever

To measure shear force the tuning fork was electrically driven using an SR-850 digital lock-in amplifier

This caused the tuning fork to oscillate parallel to the sample surface while generating an acoustic signal with

a tip-sample interaction The tuning forkrsquos electrical current was measured An acoustic sensor (R15) was

placed beneath the sample surface to measure the tip-sample acoustic signal The acoustic sensors electrical

current was measured with a second SR-850 digital lock-in amplifier Both SR-850 lock-in amplifiers passed

a dc voltage (proportional to measured signal current) to the XE-120 afm To measure normal force the XE-

120 afmrsquos laser was reflected off the tuning fork cantilever and aligned onto a position sensitive photodiode

The XE-120 performed approach and retraction curves while recording the three signals

Section 556 Performing a measurement with a Lock-in Amplifier (SR850)

An approach and retraction speed of 10 nmsec was chosen (slowest approach value) on

the XE-120 AFM To achieve a minimum sampling at every 05nm in z-displacement a

sample rate of 20hz is needed (10 nm05 nmsec=20hz) This corresponds to a maximum

time constant of Tc=10ms assuming a settling time of 5 times the time constant Tc=

1205 hz= 1s100 = 10ms To use longer time constants a slower approach speed would

be needed

Section 557 XE-120 AFM considerations

The afm was operated in nanoindentation mode which allows direct control of the

z-piezo displacement distance The tool was further operated in maintenance mode to

override insufficient laser intensity signal This was required in part because the tool has a

minimum required laser intensity to operate scanning modes and for using the automated

approach software The minimum required laser signal known as the sum peak is 11 volts

(Quadrants (A+B) values from the pspd) While utilizing a custom cantilever with a

platinum pad for increased laser signal a 0175v (A+B) laser signal was obtained

The procedure for obtaining force curves is to manually approach the sample

surface while monitoring the TF shear force signal on the digital lock-in amplifier Once

the probe encounters the surfacemesoscopic layer the tuning fork signal is observed to

decrease At this time the probe is manually retracted using software to rest above the

surface of the sample (within a 1um) The tool then is allowed to extend and retract the

probe onto and off of the surface at a constant speed of 10 nms The total distance that the

probe travels is 500nm-15um During the approach and retraction process the normal

force the shear force and the acoustic signal are obtained simultaneously

Section 558 Experimental Results

The measured normal force tuning fork amplitude and acoustic emission transducer signal

is plotted in the figures 59 510 511 below Spring constant was estimated to be

approximately 170N The zero-force point was found by averaging the pre-contact normal

force signal over approximately 300nms The zero-distance point was found by plotting

the zero normal force line until it intersected the normal force curve in the contact region

This method is described in more detail in the background section

The snap-to-contact was not observed This could be attributed to the fact that the

model used to predict a snap-to-contact is based on purely van der Waals interactions Once

a tip starts to interact with the mesoscopic water layer the model is no longer valid It

seems that the acoustic interaction occurs before the probe contacts the surface this is

concluded from the observation that the acoustic signal is observed in the region of 1-

15nm a region before the cantilever experiences a linear contact force Once the probe

enters the contact region there is a large acoustic signal that is observed It is suspected

that this is the result of the carbon tip bending as it interacts with the sample surface

Figure 59 Experimental results for Trial 1 observing normal force shear force and acoustic signal The

upper plot is the observed normal force as a function of z-displacement Notice there is not observed snap-

to-contact In the lower plot both the tuning fork amplitude (red) and the acoustic signal (blue) are plotted

with arbitrary units

top plot is the observed normal force as a function of z-displacement Notice there is not observed snap-to-

contact In the lower plot both the tuning fork amplitude (red) and the acoustic signal (blue) are plotted with

arbitrary units The acoustic signal is observed to be diminished when compared to trial 1

arbitrary units The acoustic signal is observed to be diminished when compared to trial 1 and 2

Section 559 Experimental Discussion

This experiment used a purposely stiffrdquo cantilever to avoid jump-to-contact with

the result being a cantilever which exhibits little force variance as the cantilever approaches

the sample surface Although this allows the ability to acoustically sample through the

mesoscopic region without a snap-to-contact it does not give detailed information

regarding normal force within the approach region (0-50nm) Looking at the observed

acoustic data one finds that the positioning of what is believed to be the mesoscopic water

layer is seemingly offset from the sample surface The positioning of the acoustic signal is

shifted away from the surface to a distance of 5-15nm

The stiff normal force measurements from this experiment give little in terms of

data to validate the location of the water signal which was acoustically observed An

alternate experiment was performed with a lower spring constant afm cantilever in an

attempt to confirm the location of the mesoscopic water layer relative to the surface of the

sample This experiment consisted of performing a simple approach and retraction curve

with a standard non-contact mode afm cantilever with a spring constant of approximately

40Nm ~63 humidity on a cleaved v1 grade mica surface This experiment used the

same procedure as described in section 52 for obtaining a force curve No tuning fork or

acoustic signals were gathered in this experiment only a standard approach and retraction

force curve was obtained By performing a force-displacement curve at similar humidities

(60+) but with a more sensitive (flexible) cantilever the positioning of the mesoscopic

water layer can be obtained and compared Results of the comparison between the acoustic

approach curve and the subsequent normal force curve are in agreement in that the relative

positioning of the mesoscopic water layer is offset from the surface by approximately 6-

7nm Figure 512 shows this comparison

To estimate the mesoscopic water layer thickness from the approach curve a

measurement was made at the first jump-to-contact point and also at the second jump-to-

contact point on the approach curve This is labeled in the two lower plots of figure 512

below The first jump-to-contact point is the mesoscopic layer the second jump-to-contact

is the mica surface It should be stated that for both the acoustic signal and the normal force

curve presented in figure 59-511 the same procedure to find the zero-force point (Fn=0)

and zero-surface point (Zo=0) was used This procedure is described in the background

sections 42 and 43

Figure 512 Illustrates a comparison of the relative positioning of the acoustic response obtained in current

experiment and the double snap-to-contact positioning normal force curve obtained in a verification

experiment The top two plots show the acoustic signal (blue) illustrating the offset of the acoustic signal

from the sample surface The tuning fork signal (red) (shown in the two upper plots) approaches a decreased

constant amplitude within the contact region The lower two plots show the double jump-to-contacts points

found on the normal force curve performed to verify the mesoscopic water layer position The agreement

found between the positioning of the acoustic signal and the normal force signal is of interest To be clear

the presented acoustic signal and normal force curves were obtained from two separate experiments with

separate cantilevers with separate spring constants Again the normal force curve is used here to confirm

the mesoscopic water layer positioning

Section 5510 Experimental Limitations

The experiment described in the previous section had quite a few experimental limitations

that resulted in constrained data collection The first shortfall was a low laser intensity on

the pspd We were operating a commercial afm which has a minimum required laser

intensity measured at the pspd of 11V to operate in normal conditions The custom

cantilever had a voltage of 0175v This reduced laser signal was the result of reflecting

off an irregular surface caused either by the etch manufacturing process of the tuning fork

tine or the GIS deposited platinum pad Another shortfall from using a commercial afm

was the inability to approach the sample surface at a speed less than 10 nmsec An afm

manufacturer generally has no reason to have a slower approach speed for collecting force

curve data due to sample drift at low approach speeds In our case slow approach speeds

would allow the collection of acoustic data with long time constants A long time constant

is an excellent tool to filter noise around a frequency of interest allowing the ability to

isolate the small acoustic signal

Section 56 Shear force experiments using a Commercial AFM cantilever

The objective of this experiment is to observe simultaneously the (1) afm normal

force (2) afm lateral force (C-D on the pspd) (3) tuning fork shear interaction and (4)

tuning fork acoustic signal while obtaining approach and retraction curves using a

commercial afm cantilever that has been joined with tuning fork tine Specifically this

experiment will use a commercial afm cantilever that has been electron beam platinum

welded onto a tuning fork tine The tuning fork will be oscillated parallel to the sample

surface as shown in the figure 513

Section 561 Experimental Design

A series of force curves will be obtained while the tuning fork drive voltage is

incrementally varied from 14mV to 400mV (estimated 14-40nm Tuning Fork amplitude)

The purpose is to observe how an increasing tuning fork drive voltage and thus an increase

of the shear amplitude of the afm cantilever changes the observed normal force the lateral

force the tuning fork amplitude and the tuning fork acoustic signal It is anticipated that a

larger driving voltage will result in an increased acoustic response of the afm tip interacting

in a shear motion with the mesoscopic water layer The afm is enclosed inside of an acrylic

enclosure with an air feed through line to control atmospheric humidity Humidity was

increased from ambient conditions (10-45 humidity) to approximately 60 by pumping

air at 12-15 psi through a bubbler in a beaker of deionized water Humidity was maintained

between 55-65

Figure 513 Illustrates the tuning fork oscillation direction relative to the sample surface (Image 1) The

tuning fork is driven parallel to the sample surface Image 2 illustrates the commercial afm cantilever with

no normal force applied Image 3 illustrates the cantilever with a normal force applied notice the cantilever

bending in the normal direction and the laser deflection

To observe normal force the afm laser will be reflected off of the afm cantilever

that has been welded onto a 32768 Hz tuning fork and the laser will be aligned onto the

center of the pspd Laser sum voltage (A+B) on the pspd was 135V sufficient for standard

operating conditions Vertical deflections to the afm cantilever will cause a DC voltage to

be observed in the A-B quads on the pspd figure 11 illustrates pspd operation The

geometry of the afm cantilever was measured with SEM and found to be length is 94um

width is 38um and thickness is 35-5um With Youngrsquos modulus of 130-150 GPa the

normal spring constant was found to be between 735-219 Nm For calculation of the

spring constant see section 31

lock-in amplifier (Stanford Research SR-850) The tuning fork was driven at voltages

ranging from 14-400mV While being driven the piezo-electrically generated current from

the tuning fork was observed using the same digital lock-in amplifier The drive signal was

generated and phase locked monitored using the internal frequency source of the lock-in

amplifier Sensitivity of the lock-in amplifier was set to 1uA A time constant of 3ms was

used The SR-850 digital lock-in output the magnitude (R) of the measured tuning fork

current as a DC voltage through channel 1 output This DC voltage is then passed to the

analog-to-digital-converter (ADC1) on the XE-120 afm Channel 1 has an output range

of +-10v with a maximum 10mA output The ADC1 accepts an input of +-10v with a

50kΩ impedance

Section 564 Acoustic Amplitude and Lateral Displacement Voltage Measurements

It is noted here that the commercial afm is built with 3 ADCs however nano-

indentation mode only allows recording normal force and two ADCs simultaneously To

observe 4 signals with 3 channels we needed to alternate recording a single channel

between two signals The normal force was constantly recorded The tuning fork amplitude

was always recorded on ADC1 Thus ADC2 was used to record either the acoustic

amplitude or the lateral displacement voltages in an alternating manner This means that

each experiment was performed twice in order to observe both signals for later comparison

While inconvenient to perform experiments twice it allows verification of experimental

repeatability

The commercially available acoustic emission transducers SE40-Q (Score

Dunegan SE40-Q) was used to record the acoustics generated from the shear interaction

of the afm tip and mesoscopic layer This is a different acoustic sensor than the one used

in the previous experiment Changing acoustic sensors was an attempt to increase acoustic

signal to noise ratio The SE40-Q was used as a bottom sensor and was mounted beneath

the substrate as seen in the figure 514 The acoustic signal was monitored using a second

in amplifier as a reference frequency signal The measurement was a phase locked voltage

measurement with a sensitivity of 20mV A time constant of 3ms was used The acoustic

signal was sampled similarly to the tuning fork signal in that it was passed from the channel

1 output of the lock-in amplifier into the ADC2 on the XE-120 afm ADC2 has the same

power considerations as ADC1

Section 566 Measuring the Lateral Displacement Voltage

The lateral force will be observed as a voltage value since the lateral spring constant

has not been calibrated Calibration of the lateral displacement cannot be treated as

standard since the tuning fork has a sheared oscillatory motion applied no longer is the

lateral displacement purely a result of lateral friction forces but it includes a driven

oscillatory shear component In our experiment the lateral displacement is proportional to

the laser deflection on the C-D quadrants on the pspd ie voltage signal results from

horizontal displacements of the laser spot on the pspd The lateral displacement voltage is

recorded by passing the raw lateral force voltage component from the pspd out to a second

SR850 lock-in amplifier using a signal access module (SAM) to share the voltage between

the afm electronics and the lock-in amplifier The lock-in amplifier is used to track the

shear lateral oscillation that results from the motion of the oscillating tuning fork A 3ms

time constant and a sensitivity was 1V was used by the lock-in amplifier The magnitude

(R) of the lateral displacement voltage was passed from the channel 1 output of the lock-

in amplifier into the ADC2 on the XE-120 afm

A cleaved mica (PELCO Mica Disc 99mm Grade V1) surface was used The mica

was cleaved using the scotch-tape method such that a single layer of mica is peeled away

revealing a pristine surface free of non-native surface contaminants The probe tip was

aligned near the edge of the mica sample (within 50um of the edge) This was an attempt

to avoid unwanted contact of the tuning fork and the substrate It is only desired that the

probe tip and substrate have interaction The sample was adhered to the acoustic emission

sensor using vacuum grease (Dow Corning High Vacuum Grease)

The two schematic figures below illustrate the two separate experimental procedures

needed to measure the 4 signals of interest Figure 514 illustrates the experimental

procedure for measurement of normal force tuning fork amplitude and acoustic amplitude

Figure 515 illustrates the illustrates the experimental procedure for measurement of

normal force tuning fork amplitude and lateral displacement voltage

Figure 514 This schematic shows the basics of the experiment where normal force tuning fork amplitude

and acoustic amplitude are measured The laser is aligned on the afm cantilever that has been welded onto a

tuning fork The tuning fork is driven at open air resonance with displacement being parallel to the sample

surface A mica surface is used The two lock-in amplifiers measure the tuning fork amplitude and acoustic

amplitude These measurements were passed to the afm on adc1 and adc2 of the XE-120 afm control

electronics The XE-120 afm also recorded the normal force-displacement curve

Figure 515 The schematic shows the basics of the experiment where normal force tuning fork amplitude

and lateral displacement voltages are measured The PSPD records voltages proportional to vertical and

lateral laser displacements The XE-120 afm records the vertical laser deflections A signal access module is

used to output the lateral displacement voltage to lock-in amplifier 2 The two lock-in amplifiers measure

the tuning fork amplitude and lateral displacement voltages respectively These measurements are passed to

the afm on adc1 and adc2 of the XE-120 afm control electronics The XE-120 afm also recorded the normal

force-displacement curve

An afm cantilever approach and retraction speed of 10 nmsec was chosen (slowest

approach velocity) on the XE-120 AFM To achieve a minimum sampling at every 05nm

in z-displacement a sample rate of 20hz is needed (10nm 05nmsec = 20hz) This

corresponds to a maximum time constant of Tc=10ms assuming a settling time of 5 times

the time constant Tc= 1(205) hz= 1sec100 = 10ms To use longer time constants a

slower approach speed would be needed or concessions on the Z-axis sampling resolution

We use a 3ms time constant for all lock-in measurements

The afm was operated in nanoindentation mode which allows direct control of the z-piezo

displacement distance The procedure for obtaining force curves is to bring the afm

cantilever to a distance of approximately 100um above the sample surface The software

performed the final approach automatically engaging the sample surface and maintaining

a fixed force on the sample surface at a set point of 350nN The tool then is allowed to

extend and retract the probe onto and off of the surface at a constant speed of 10 nms The

total distance that the probe travels is 500nm-15um During the approach and retraction

process the normal force the shear force and the acoustic signal or lateral force are

obtained After each approach and retraction curve is obtained the probe is removed from

the surface and the surface is re-approached for each subsequent curve collection in an

attempt to reduce probe wear

Two sets of data are plotted below The first data set shows the normal force tuning fork

amplitude and the acoustic amplitude The second data set shows the normal force tuning

fork amplitude and the lateral displacement voltage General trends will be discussed

Humidity was approximately 55-60

Data Set 1

Plots show the normal force tuning fork amplitude and the acoustic amplitude

1 Normal Force Figure 516 Figure 517

2 Tuning Fork Amplitude Figure 518 Figure 519

3 Acoustic Amplitude Figure 520

4 Low Amplitude (50mV) and High Amplitude (300mV) Comparison

a Figure 521- Low tuning fork drive voltage

b Figure 522- High tuning fork drive voltage

Figure 516 Data-Set1 A series of normal Force vs Z-Displacement approach curves plotted as tuning

fork drive voltage is increased from 50mV to 400mV Tuning fork (TF) drive voltage is labeled in the top

right of each plot Notice a large force builds near the sample surface as the tuning fork drive voltage is

increased Notice that a low tuning fork drive voltage has a corresponding force curve that resembles a

standard force curve with two snap-to-contact points The two snap-to-contact points are representative of

the beginning and the end of the mesoscopic layer Notice at low TF drive voltages (50-100mV) that the

normal force curve is attractive before surface contact and at large TF amplitudes (gt100mV) the normal force

is repulsive

Figure 517 Data-Set1 The series of Normal Force vs Z-Displacement approach curves plotted

overlapping as tuning fork drive voltage is increased from 50mV to 400mV Tuning fork drive voltages are

labeled in the top right of the plot Notice a large force builds near the sample surface and expands further

from the sample surface as the tuning fork drive voltage is increased

Figure 518 Data-Set1 A series of Tuning Fork Amplitude vs Z-Displacement approach curves plotted as

tuning fork drive voltage is increased from 50mV to 400mV Tuning fork drive voltage is labeled in the lower

right of each plot The most interesting trend observed is that initially there is a large temporary tuning fork

amplitude decrease near the sample surface after which the amplitude rebounds and continues to decrease

linearly This dip in the tuning fork amplitude corresponds spatially exactly in the sample region as the normal

force bump

Figure 519 Data-Set1 The series of Tuning Fork Amplitude vs Z-Displacement approach curves are

overlapped to illustrate the general trends of increased tuning fork amplitude damping Notice the amplitude

dependency of the dampening distance ie the distance Δz at which tuning fork signal is attenuated is directly

dependent upon initial tuning fork amplitude This was observed in non-linear force models proposed by MJ

Gregor et al in (1995)

Figure 520 Data-Set1 A series of Acoustic Amplitude vs Z-Displacement approach curves plotted as

right of each plot At low tuning fork driving voltages 50mV-116mV there is little acoustic response as the

afm cantilever is interacted with the mesoscopic water layer near the sample surface As the tuning fork

amplitude is increased above 130mV an acoustic response is observed The acoustic signal shows an increase

in magnitude between the sample surface and 25nm This region corresponds to an increase in observed

normal force and a decrease in tuning fork amplitude as described in the previous plots

Figure 521 A low amplitude tuning fork approach curve with distinct interaction regions labeled Region

(A) is the complete contact region (B) is the lsquoknockingrsquo contact region (C) is the 3rd body contact region

(D) is the Van der Waals region and (E) is the long range non-contact region Top plot is the normal force

curve middle plot is the tuning fork amplitude and the bottom plot is the observed acoustic signal The

drive voltage of the tuning fork was 50mv or approximately an estimated 5nm drive amplitude Region B

spans approximately 16nm region C spans 16nm The Van der Waals region D was extended arbitrarily to

the 50nm region but the forces extend weakly beyond Labeled point (1) is the kink point point (2) is the

sample surface point (3) is the initial contact point with the mesoscopic layer

Figure 522 A high amplitude tuning fork approach curve with distinct interaction regions labeled The drive

voltage of the tuning fork was 300mv or approximately an estimated 30nm drive amplitude Region (A) is

the complete contact region (B) is the lsquoknockingrsquo contact region (C) is the 3rd body contact region (D) is a

weakly repulsive region and (E) is the long range non-contact region Top plot is the normal force curve

middle plot is the tuning fork amplitude and the bottom plot is the observed acoustic signal Region B spans

approximately 297nm region C spans 16nm The region D here does not follow the attractive van der waals

behavior this regions weakly repulsive region extends about 50nms Labeled point (1) is the kink point point

(2) is the sample surface point (3) is the contact point with the mesoscopic layer point (4) is the beginning

of the repulsive interaction Notice the well-defined acoustic response in region C

Data Set 2

Plots shows the normal force tuning fork amplitude and the lateral displacement

3 Lateral Displacement Figure 527 Figure 528

a Figure 529-Low tuning fork drive voltage

b Figure 530-High tuning fork drive voltage

fork drive voltage is increased from 14mV to 350mV Tuning fork drive voltage is labeled in the toplower

the beginning and the end of the mesoscopic layer Notice that at low shear amplitudes (14-74mV) the normal

force is attractive and at high amplitudes (gt74mV) the normal force is purely repulsive

Figure 524 Data-Set2 The series of Force vs Z-Displacement approach curves plotted overlapping as

tuning fork drive voltage is increased from 14mV to 350mV Tuning fork drive voltages are labeled in the

upper right of the plot Notice a large force builds near the sample surface and expands further from the

sample surface as the tuning fork drive voltage is increased

right of each plot Similar to data set 1 the tuning fork signal exhibits a dampened bump near the sample

surface

overlapped to illustrate general trends of increased tuning fork amplitude damping Dampening occurs further

from the sample surface as tuning fork drive voltage is increased This corresponds to the increased normal

force bump observed further from the sample surface

Figure 527 Data-Set2 A series of Lateral Displacement vs Z-Displacement approach curves plotted as

tuning fork drive voltage is increased from 14mV to 350mV Tuning fork drive voltage is labeled in the upper

right of each plot Two interesting trends are observed in the lateral force approach curves Initially at low

tuning fork drive voltages (14mV-74mV) the lateral force is constant while the afm tip is above the sample

surface and increases to a larger constant value once tip-surface contact is established As the tuning fork

drive voltages are increased above 100mV the lateral displacement rapidly decreases once the afm tip enters

the mesoscopic region (0-25nm) and increases again once surface contact occurs The region of rapid

decrease in lateral displacement signal corresponds directly to the observed increase in normal force the

decrease in tuning fork amplitude and an increase in acoustic response

Figure 528 Data-Set2 The series of Lateral Displacement vs Z-Displacement approach curves are

overlapped to illustrate the increase in lateral displacement dampening as the tuning fork drive voltage is

increased Tuning fork drive voltage is labeled in the lower right Lateral displacement is proportional to

friction forces and it is observed that during tip-sample contact a large friction force is present All curves

are aligned to the dampened axis as the zero point with the assumption that the lateral displacement is

constant in this region regardless of initial TF drive voltage

(D) is the van der waals region and (E) is the long range non-contact region Top plot is the normal force

curve middle plot is the tuning fork amplitude and the bottom plot is the observed lateral displacement

signal The drive voltage of the tuning fork was 25mv or approximately a 25nm drive amplitude Region B

spans approximately 7nm region C spans 83nm The van der waals region D was extended arbitrarily to the

50nm region but the forces extend weakly beyond Labeled point (1) is the kink point (2) is the sample

surface point (3) is the 3rd body initial contact point

Figure 530 A high amplitude tuning fork approach curve with distinct interaction regions labeled Region

spans approximately 302nm region C spans 258nm The region D here does not follow the attractive van

der Waals behavior this regionrsquos weakly repulsive region extends about 50nms Labeled point (1) is the kink

point (2) is the sample surface point (3) is the 3rd body initial contact point

Figure 531 Illustrates the observed lsquokinkrsquo in the normal force curve and the static lsquoclampingrsquo effect for the

tuning fork amplitude The lsquokinkrsquo and static lsquoclampingrsquo occur simultaneously at the transition between the

complete contact region A and the lsquoknockingrsquo region B The tuning fork drive amplitude was 350mV ~35nm

amplitude

Section 5611 A lsquoKinkrsquo in the Normal Force Curve

The normal force approach curve exhibits a kink figure 531 which is described as

a sudden change in slope between regions A amp B In literature3 this kink is the point where

mechanical contact with the probe and sample surface is initiated

Section 5612 A lsquoStatic Clampingrsquo of the Tuning Fork

Static clamping is the mechanical clamping of the probe to the sample surface At

this point in the tuning fork approach curve shown above the tuning fork amplitude

spontaneously decreases to a minimum This is attributed to micro-surface roughness of

the sample and probe3

Section 5613 Estimated Tip Velocities

Table 18 Tuning Fork Shear Velocity (Estimated)

Tuning Fork Drive Voltage Tuning Fork Displacement

Shear Velocity (Vel)

14mv 14nm 288 umsec

25mv 25nm 514 umsec

50mv 50nm 1030 umsec

Relation for Tuning fork Drive Voltage and Tuning fork Displacement ~10mvnm

Tuning fork Drive Frequency FResonance = 32801 Hz Tuning fork Displacement = A0

W = 2πFResonance

Vel = -A0wsin(wt)

Velocity equation obtained by differentiating Simple Harmonic Oscillator (SHO) position

equation

Section 5614 Region A Complete Contact Region

Region A is accepted as the direct mechanical contact between the tip and substrate

Normal force is observed to follow Hookersquos law and linearly increase with z-piezo

displacement The tuning fork amplitude is dampened to a constant amplitude The acoustic

signal in this region is observed to be constant in amplitude The observed lateral

displacement signal is a maximum in this region A constant but maximum lateral

displacement could be indicative of a tip in frictional shear motion with the surface

possibly indicating that the tip maintains a constant dampened shear motion while direct

contact occurs as has been observed by previous experiments5

Section 5615 Region B The lsquoKnockingrsquo and Partial Contact Region

Region B corresponds to a region where the tip is in an increasingly dampened

motion as probe height is decreased resulting from a lsquoknockingrsquo which occurs between the

tip and sample surface The lsquoknockingrsquo term is used describe the tip tapping the surface

At the transition between regions A and B there occurs a kink in the normal force curve

as demonstrated in figure 531 This kink is observed as a change in slope of the force curve

into a constant Hookersquos law regime A previous study by MJ Gregor et al describes this

kink in the force curve as being indicative of the initial tip-sample contact3 This is in subtle

disagreement to what we observe We observe a dampening of the tuning fork amplitude

occurring throughout region B indicative of a probe tapping the sample surface and at the

kink point a firm mechanical contact is established between the probe and surface

resulting in a completely dampened tuning fork amplitude and a linear force Once again

referencing Gregor et al within a nanometer of the kink region the tuning fork was shown

to exhibit a spontaneous jump-to-zero amplitude a point at which it is said that a lsquostatic

clampingrsquo of the probe occurs caused be the lsquomicro-roughnessrsquo of the sample surface3

The acoustic response in region B mirrors that of the tuning fork amplitude

indicating that the tuning fork remains in motion The lateral displacement is observed to

increase in magnitude as tip-sample lsquoknockingrsquo increases in intensity This corresponds to

a probe encountering a shear dampening indicative of a cantilever exhibiting increased

torsional and lateral motion from lateral friction interactions

Section 5616 Region C The 3rd Body Contact Region

The primary question of interest of this thesis is what is occurring in region C a

region defined roughly from the surface z=0 to approximately 15-25 nm from the surface

It is best to examine this region under two separate conditions a low tuning fork drive

amplitude (14-50mV) and a large tuning fork drive amplitude (300mV)

At low tuning fork drive amplitudes (drive voltage 14-50mV estimated shear

amplitude of 1-5nm) the normal force curve exhibits the typical attractive forces and

double snap-to-contact points characteristic for the presence of a water layer Figure 529

It is observed that the normal force in this region is negative and attractive Such attractive

conditions are not characteristic of a double layer repulsion effect The tuning fork

amplitude exhibits a dampening with a slope different from the lsquoknockingrsquo of region B and

the dampened region A possibly indicative of a different medium supplying the dampening

interaction The is no acoustic response observed at these low drive amplitudes The lateral

displacement is observed to decrease in this region indicating a low shear interaction or

correspondingly a decrease in tip shear motion

At increasingly higher tuning fork drive amplitudes (drive voltage 100-300mV

estimated shear amplitude 10-30nm) the normal force is observed to increase and

transitions from a net attractive force to a net repulsive force Figure 530 Also a large

lsquobumprsquo appears in the same region as the mesoscopic water layer that was previously

observed at low tuning fork drive amplitudes This lsquobumprsquo in the force curve has observed

force values approaching 1500nN at large shear amplitudes This bump has a peak at

approximately the midpoint of region C and itrsquos exact position is seemingly dependent

upon the tuning fork drive amplitude After the peak the normal force is observed to

decrease until a lsquoknockingrsquo contact with the surface is established at the point between

region B and C Figure 530 The tuning fork amplitude is observed to decrease in the

sample spatial region that the increase in the normal force occurs A local minimum a dip

in the tuning fork amplitude is observed to occur at the same point as the peak in the normal

force After this tuning fork dip is observed the tuning fork recovers to a less dampened

state The acoustic response in this region is observed to follow that of the normal force

As the probe moves through region C an increase in acoustic signal is found to correspond

to a decrease in tuning fork amplitude possibly indicative of a mechanical dampening

interaction resulting in the generation of an acoustic response The lateral displacement is

observed to decrease and exhibits a minimum corresponding spatially to the tuning fork

amplitude minimum The observed lateral displacement is expected to decrease in relation

to a decrease in tuning fork amplitude with small friction force or increase in the presence

of a large frictional force It appears that the former condition is present in this region ie

a dampened tuning fork motion in the presence of a small friction force provided by the

probe-mesoscopic layer interaction

Section 5617 Region D The Mid-Range Interaction Region

At low drive amplitudes (14-50mV) this region exhibits the typical behavior of van

der Waals forces as observed in the normal force plots of figure 529 At high drive

amplitudes (300mV) this region exhibits a long range repulsive force figure 530 Long

range repulsion is typical of electronic repulsion Section 5620 speculates sources for this

observed transition from attractive to repulsive forces

Section 5618 Region E The Long Range Non-Interaction Region

This is loosely defined as the far region where van der Waals forces or repulsive

forces are below thermal background force levels

Section 5619 Plausible Explanations for the Observed lsquoBumprsquo in Region C

It has been shown in previous experiments by Peter M Hoffmann et al37 that a

sharp transition occurs from viscous (fluid like) to elastic (solid like) response of a confined

liquid which is highly dependent upon vertical probe compression rates Hoffmann termed

this material response dynamic solidification observing a high elastic response in the

atomically ordered water layers closest to the sample surface Hoffmann found that at a

vertical approach velocity of 8 Angstroms per second the remaining two layers of water

exhibited a large elastic response and in addition finding that at a vertical approach

velocity of 14 Angstroms per second the remaining 4 water layers exhibited a large elastic

response In essence Hoffmann illustrated a dynamic solidification where the viscous and

elastic response was highly dependent upon vertical probe approach velocity He illustrated

a threshold vertical compression rate of 8 angstroms per second showing that elastic

behavior is exhibited when compression rates are greater than this and a viscous response

at compression rates below this threshold Hoffmann interpreted the large elastic response

as the water layers exhibiting a solid-like behavior In our experiment the probersquos vertical

compression rate was approximately 7 times larger than Hoffmannrsquos largest at 100

Angstroms per second or 10nms At these rapid compression rates it is plausible that the

ordered liquid layers could exhibit a high elastic stiffness

In our experiment in addition to high vertically applied compressive loads we

applied a shear interaction at various velocities Table 18 shows the estimated probe

velocities given as a function for tuning fork drive amplitudes In estimating shear velocity

we differentiated a simple harmonic oscillator (SHO) to establish a velocity relation We

found that as the shear velocity increased above a threshold the observed interaction

exhibited an elastic response characteristic of a compressed solid-like material Under our

experimental conditions it is estimated that the threshold velocity for a shear dynamic

solidification occurs at tuning fork drive voltage of 116mv or an estimated tuning fork

amplitude of 116nm This corresponds to an estimated probe shear velocity of 2389 umsec

from table 18 Though a high uncertainty exists in the actual shear displacement and

velocity at which the threshold for which this transition occurs qualitatively it is shown

that there exists a distinct transition between a viscous water medium at low shear

amplitudes and an elastic solid-like medium at high shear amplitudes

Evidence for an elastic response of the 3rd body or water layer observed in region

C is evident when analyzing the normal force and the acoustic response data At high shear

amplitudes the normal force was observed to have a large bump that can be interpreted as

the probe lsquocompressingrsquo a solid-like material Correspondingly the acoustic response in

shows an increase in amplitude which could be an elastic interaction between the probe

and the 3rd body in region C An elastic interaction between the probe and the 3rd body

would exhibit a subsequent dampening of the tuning fork amplitude which is observed in

this region

Interestingly as the probe is moved through the 3rd body towards the surface the

dampening is gradually decreased until the transition from region C to B occurs At this

point the tuning fork amplitude is no longer dampened and the lsquobumprsquo in the normal force

has disappeared One possible interpretation of this observation could be that the 3rd body

medium is lsquosqueezed outrsquo between the probe and the sample surface Our data illustrates

that the dampening subsides at the transition from region C into B Recall region B

corresponds to the tip making mechanical knocking-contact with the sample surface

leaving scant space for a 3rd body This supports the possible squeezing out of the 3rd body

medium

Section 5620 Transition from Attractive to Repulsive Response

Examining the normal force response as the tuning fork amplitude is increased

figure 523 it is observed that the normal force is attractive in nature in the pre-contact

regions C amp D while operating at low shear amplitudes However as the tuning fork drive

voltage was increased the observed normal force transitions into purely a repulsive

response On possible explanation for this response entails the phase response of the tuning

fork oscillator as a function of tuning fork drive amplitude In studies on the amplitude and

phase response of tapping mode afm 38 39 it was observed that ldquoif the free oscillation

amplitude exceeds a critical threshold that the amplitude-distance and phase-distance

curves exhibit a distinct transition from a net attractive force between the tip and sample

to a net repulsive responserdquo 39 Zitzler reasoned this effect ldquocan be explained by assuming

the intermittent formation and rupture of a capillary neck in each oscillation cycle of the

AFM cantileverrdquo

In our experiment we attempted to drive the cantilever parallel to the sample

surface however any misalignment results in a vertical component in the oscillation

vector Thus any increase in the shear driving voltage could result in a similar tapping of

the mesoscopic layer effect observed in the tapping mode experiments by Garcia Paulo

and Zitzler with the probe tapping in the normal direction

Chapter 6 Future Experiments

The use of acoustics for sample characterization on the nanoscale is becoming a mature

technology It is useful to consider possible uses for acoustics for semiconductor

characterization of surface roughness defects and void identification In addition to

characterization investigation of surface manipulation with acoustics should be explored

Section 61 Future Force Curve Experiments

A couple experiments that would clarify the nature of shear force microscopy are

Compare the Amplitude-Frequency response (Frequency Sweep) as function of

distance above a sample Previous work by Karrai 1 and Gregor 3 exhibit

inconsistencies resulting from both experiments operating at two drastically

different amplitudes regimes with Karrai observing a simple harmonic oscillator

response and Gregor observing a non-linear response By performing frequency

sweeps at small tuning fork amplitudes(lt1nm) and increasing the amplitude to

large tuning fork amplitudes (gt30nm) in small steps it is possible to illustrate both

parties are correct within their respective experimental domains This would

illustrate that the shear force is a result of a knocking between the probe and sample

surface or 3rd body medium

Another series of experiments to perform in the future would be a repeat our

previous experiment described in section 56 while also capturing the phase

response as the probe moves through the bump region Zitzlerrsquos tapping mode

experiment predicts a negative change in phase response when the probe is in the

repulsive force region By investigating the probe while it is interaction in the

bump region C it is possible to isolate the role that phase plays in the cantilever

dynamics

Section 62 Future Experiment 1 Subsurface void localization

Section 621 Proposal

It has been shown in previous work that it is possible to use nearfield acoustics to localize

subsurface voids within many materials However most previous work has only explored

the localization of either large subsurface voids (10rsquos of um2) located under shallow

substrate or nanoscale defects (~1nm2) under atomically thin substrate such as graphene

There is a need to develop subsurface acoustic imaging across a spectrum of void sizes and

void depths Further the investigation of alternative substrate materials is warranted ie

explore a Cu Oxide substrates as well as many different substrate materials which are

used in semiconductor device fabrication

Section 622 Implementation

Subsurface void localization has been implemented using the XE-120 afm A sample was

mounted on top of a vertically oscillating piezo-electric The sample and piezoelectric are

driven together to oscillate at resonance (about 70-80 kHz) using a SR850 lock-in amplifier

with drive voltages ranging from 50 mV to 500 mV A contact mode afm cantilever of low

spring constant (lt05nNnm) is scanned across the sample to obtain an image The afm

cantilever will have a vertical deflection oscillation component that is unique to the

Youngrsquos modulus of the sample substrates chemical composition and geometry The lock-

in amplifier is used to measure the vertical deflection of the afm laser spot (A-B) on the

pspd The laser signal is passed from the signal access module to the lock-in where both

the amplitude and phase of the laser signal is tracked Output channels 1 and 2 of the lock-

in amplifier are used to pass the DC components to the afm control and scan electronics

where it is plotted as the afm cantilever is raster scanner across the sample surface It is

found that a slow afm scan speed (02Hz) give a well-defined ultrasonic amplitude and

phase signals

Section 623 Sample Fabrication

A geometrically known subsurface feature was fabricated using electronion-beam

deposition and milling This made use of a dual-beam FEI Helios 400s with gas injection

systems for platinum and carbon deposition On a freshly cleaved silicon substrate a series

of box outlines of decreasing sizes were deposited andor milled onto the surface The box

outline sizes were 500x500 nm2 1x1 um2 2x2 um2 3x3 um2 4x4 um2 5x5 um2 Three

sets of boxes were fabricated Set 1 was fabricated using electron-beam platinum

deposition using a low resolution electron beam at 2kV 340pA Set 2 was fabricated using

high resolution electron beam platinum deposition 2kV 340pA The difference between

low resolution electron beam deposition and high resolution deposition is the use of an in-

lens secondary electron detector for high resolution imaging which has far higher

resolution due to field enhancement that channels electrons to the detector Low resolution

imaging is accomplished using in chamber detectors which typically utilize a bias voltage

to filter and collect electrons of various energies Set 3 was fabricated using ion beam

milling at 30kV 300pA Once the three sets of box outlines were fabricated they were

covered with ion beam deposited carbon Ion beam conditions were 30kV 1nA The

fabrication of the box outlines is shown in the figure 61 below The carbon capping of the

features is shown in the figure 62 below

Figure 61 Demonstrates electron and

ion beam fabrication of subsurface

features Green boxes illustrate the box

outlines that were deposited using low

resolution electron beam platinum

deposition Grey boxes illustrate the box

outlines that were deposited using high

deposition Yellow boxes illustrate the

box outlines that were fabricated using

ion beam milling Box outlines were

organized in order of increasing size and

the 3 sets were fabricated within a 30um2

Figure 62 This illustrates the ion beam deposited carbon capping

of the fabricated box outlines The upper diagram shows the

carbon capping of the first set of box outlines The grey outline is

used to signify a carbon deposition The middle diagram shows

the center set of box outlines being carbon capped Notice the

overlapping depositions this was an attempt to minimize changes

in topographic height The lower diagram illustrates the carbon

capping of the final set of box outlines

Figure 63 Shows a schematic outline of the fabricated features that have been buried under an ion beam

deposited carbon layer (Image A) The dotted red outline is used to show the afm scan orientation with

respect to the features Image B is a topographic afm scan of the fabricated features Notice that the carbon

layers have a small height difference at the overlap region Also the low-resolution electron beam features

are visible

Figure 64 Image illustrates a measurement of the carbon cap height of approximately 500nm (Left) (Right)

image illustrates a measurement of the 3x3 box outline It is found to be 285 um wide

Figure 65 A 3D topographic afm image of the scanned features Low resolution electron beam deposited

(LrEBD) features are the foremost features and the middle region is the high-resolution electron beam

deposited (HrEBD) feature set and the back set of features are the IBD features This image is used to

illustrate the ion beam deposited capping capping overlap and surface height of the subsurface feature

Notice that LrEBD features are easily visible on the surface topography while HrEBD and IBD features are

not visible

Figure 66 Shows a schematic outline of the fabricated features that have been buried under a ion beam

deposited carbon layer (Above) Image A The dotted red outline is used to show the afm scan orientation

with respect to the features Image B is a topographic afm scan of the fabricated features Image C is a

normal force image Image D is a lateral force image Notice the presence of the fabricated features vertical

lines from the ion beam carbon deposition and regions of high lateral force from the afm tip dragging surface

debris as the image is scanned Image E is the ultrasonic amplitude image obtained by lock-in tracking the

vertical laser deflection Subsurface features are identifiable in all three regions low resolution EBD high

resolution EBD and ion milled box outlines Image F is the ultrasonic phase image obtained by lock-in

tracking the vertical laser deflection Again components of all sets of features are identifiable

Figure 67 (Above) The same images as the previous figure but with the subsurface features labeled with

red dotted outlines

The goal was to fabricate features that were predominantly subsurface to verify that the

UAFM technique is sensitive to subsurface defects It was found that the fabrication

process resulted in small but identifiable surface features that were identified using

topographic afm It was found that subsurface features are identifiable using the UAFM

technique specifically the by monitoring the amplitude and phase

Section 625 AFM Topography

In the observed topographic image Figure 67 Image B the vertical set of rectangles to the

left is the low resolution EBD features the middle set was the high resolution EBD features

and the set to the right was the ion milled features In the topographic image only the low

resolution EBD features were easily identifiable

Section 616 AFM Normal Force Image

The normal force image Figure 67 Image C is used as the error signal while scanning in

contact mode therefore it is expected to have a low contrast and low sensitivity to surface

features It does exhibit high contrast at the transition from silicon substrate to IBD carbon

capping pad and once more at the overlap region between capping pads identified as

bright vertical lines in image C This was likely caused by the gain parameter being too

low All three sets of box outlines are visible to a varying degree Similar to the topographic

image the low resolution electron beam deposited LrEBD features are easiest to identify

while HrEBD and IBD features are less defined

Section 617 AFM Lateral Force Image

The LFM image Figure 67 Image D is sensitive to lateral forces that induce a cantilever

torsion In image D material changes height changes produce image contrast In the image

LFM there exists a ldquodebris trailrdquo that resulted from a piece of surface debris being dragged

across the IBD capping pad and is visible as a white streak across the image The high

contrast of this debris trail indicates that it produces a large friction force on the cantilever

as it is contact mode scanned across the sample surface This feature is not visible in the

topographic image The three sets of subsurface features are visible to varying degrees All

features for LrEBD boxes are visible while 1x1um2 features or larger are visible in the

HrEBD and IBD feature sets Feature visibility is summarized in the table 19

Section 628 UFM Amplitude Image

The UFM amplitude image figure 67-Image D is sensitive to changes in material

properties such as Youngrsquos modulus and surface topography Image E has high contrast

regions in areas exhibiting large changes in surface height (Overlapping IBD capping

regions) in the region containing the ldquodebris trailrdquo and in regions containing subsurface

features Similar subsurface feature visibility to the LFM images is noted Submicron

features are not well defined Feature visibility is summarized in the table 19

Section 629 UFM Phase Image

The UFM phase image is sensitive to phase changes that are induced as the afm cantilever

and vertically oscillating surface interact Feature visibility is similar to LFM and UFM

amplitude Feature visibility is summarized in the table 19

Table 19 Summary of Feature Visibility

Deposition

500x500nm2

1x1 um2 2x2 um2 3x3 um2 4x4 um2 5x5 um2

Topography LrEBD X X X X X X

Normal Force LrEBD X X X X X X

Lateral Force LrEBD X X X X X X

Amplitude

LrEBD X X X X X X

UFM Phase LrEBD X X X X X X

X = Visible Feature = Partially Visible Feature Blank = non-Visible Feature

Section 63 Future Experiment 2 Defect Adhesion

Through the analysis of the afm tip-substrate interaction it was realized that adhesion

energy is a function of surface area of interaction In the semiconductor industry the

presence of particle defects can cause device failure and reduce production yields By

exploring particle size distributions methods can be developed for particle removal or

particle adhesion mitigation In ambient conditions the major contributor for particle

adhesion is due to capillary forces due to high humidity In low vacuum conditions such as

those used in atomic layer deposition (ALD) plasma enhanced chemical vapor deposition

capillary forces are reduced and the major interaction force of non-charged particles is van

der waals Controlling humidity could have a large effect on defect numbers

Section 631 Defect Prevention

By directly measuring adhesion forces as a function of surface area one can estimate

binding energies and therefore the work required to remove defect debris It is theorized

that by acoustically exciting a sample surface or wafer at a mechanical resonance that

defect adhesion energy can be overcome Perhaps by acoustically exciting a sample during

processing defects can reduced or directed with acoustic nodes to controlled sacrificial

regions Perhaps wafers could also be acoustically cleaned reducing process time and cost

Please note the method described here is not ultrasonic bath cleaning which works on the

basis of the formation of air bubbles that supersonically collapse to apply force to particle

debris Such method can damage 1um2 areas of substrate and are not ideal for the current

submicron feature The process described above would make use of passing acoustic waves

through the wafer which then are transferred in the normal direction onto surface defects

in both ambient and vacuum conditions

Figure 68 Schematic illustrating using an acoustic sensor-transducer to apply a force to the surface of a

wafer with surface debris present Such acoustic cleaning could transfer an acoustic signal through the wafer

using direct mechanical contact allowing the process to be applied in both ambient and vacuum conditions

Section 64 Surface Nano-Roughness Characterization using Acoustics

In the semiconductor field surface roughness analysis is typically performed to

characterize resistivity of a nanofilm and verify uniformity of Atomic Layer Deposition

(ALD) processing A surface roughness analysis is performed using an afm and the results

are dependent upon tip radius As an afm tip is used to process multiple samples the tip

becomes blunted and loses its ability to characterize small cracks and voids in a sample

surface Afm tips can be expensive and it is time consuming to change tips after multiple

samples A method that is able to reliably measure surface roughness without contact is

needed It is proposed here that the relationship between nanosub-nano roughness and

surface acoustic reflectivity be investigated to see if a useful correlation is present It has

been shown by Nagy and Alder40 that the acoustic reflection from a sample surface is

significantly attenuated by a rough surface It has been shown that the surface roughness

changes of 3-4nm of a thin-film resonator results in a drastic reduction in the Q-factor from

350 down to 15041 This suggests that acoustic reflectivity losses are sensitive to nanometer

roughness to changes

A tuning fork driven to oscillate at resonance is mounted a fixed distance above a sample

surface and is driven such that motion is perpendicular or oblique (angled) to the sample

surface An acoustic emission transducer (Acoustic Sensor) is mounted at a fixed angle and

distance from the sample and tuning fork The experiment could utilize a simple glass slide

half coated with metal deposition and half without a metal deposition as a test surface

AFM and SEM imaging can be used for direct roughness measurements and topographic

surface analysis

Figure 69 (A) Illustrates an acoustic wave is incident upon a smooth surface that exhibits specular

reflection (B) An acoustic wave is incident upon a rough surface and the acoustic signal is diffusely scattered

Section 643 Estimated Results

It is expected that there will be a relationship between surface acoustic reflectivity and the

nano-surface roughness However the magnitude and usefulness of this relation remains

unknown

References

1 Karrai K amp Tiemann I Interfacial shear force microscopy Phys Rev B Condens

Matter Mater Phys 62 13174ndash13181 (2000)

2 Eaton P amp West P Atomic Force Microscopy (2010)

3 Gregor M J Blome P G Schoumlfer J amp Ulbrich R G Probe‐surface interaction

in near‐field optical microscopy The nonlinear bending force mechanism Appl

Phys Lett 68 307ndash309 (1996)

4 Hsu K amp Gheber L A Tip-sample interaction in a lsquoshear-forcersquo near-field

scanning optical microscope Rev Sci Instrum 70 3609ndash3613 (1999)

5 Okajima T amp Hirotsu S Study of Probe-Surface Interaction in Shear-Force

Microscopy Effects of Humidity and Lateral Spring Constant Opt Rev 5 303ndash309

(1998)

6 Fernandez Rodriguez R amp Rodriguez R F Confined Mesoscopic Fluid-like Films

Analyzed with Frequency Modulation and Acoustic Detection (2000)

doi1015760etd2046

7 Li T-D Atomic force microscopy study of nano-Confined liquids (Georgia

Institute of Technology 2008)

8 Chau A Regnier S Delchambre A amp Lambert P Influence of geometrical

parameters on capillary forces in 2007 IEEE International Symposium on Assembly

and Manufacturing (2007) doi101109isam20074288475

9 Sparks D L Environmental Surfaces and Interfaces from the Nanoscale to the

Global Scale J Environ Qual 39 1535 (2010)

10 Hu J -D Xiao X Ogletree D F amp Salmeron M Imaging the Condensation and

Evaporation of Molecularly Thin Films of Water with Nanometer Resolution

Science 268 267ndash269 (1995)

11 Advincula R C Nano-Surface Chemistry Edited by Morton Rosoff (Long Island

University) Marcel Dekker Inc New York and Basel 2002 xii 678 pp ISBN 0-

8247-0254-9 J Am Chem Soc 124 12630ndash12631 (2002)

12 Lemoine P Roy S S Quinn J P Maguire P D amp McLaughlin J A D Carbon

nanostructures grown with electron and ion beam methods Appl Phys A Mater

Sci Process 86 451ndash456 (2007)

13 Hamaker H C The Londonmdashvan der Waals attraction between spherical particles

Physica 4 1058ndash1072 (1937)

14 Argento C amp French R H Parametric tip model and forcendashdistance relation for

Hamaker constant determination from atomic force microscopy J Appl Phys 80

6081ndash6090 (1996)

15 Atomic Force Microscopy Peter Eaton and Paul West MRS Bull 39 379 (2014)

16 Cappella B amp Dietler G Force-distance curves by atomic force microscopy Surf

Sci Rep 34 1ndash104 (1999)

17 Das S Sreeram P A amp Raychaudhuri A K A method to quantitatively evaluate

the Hamaker constant using the jump-into-contact effect in atomic force

microscopy Nanotechnology 18 035501 (2007)

18 Atomic Force Microscope Modes amp Technique | Park Systems Available at

httpwwwparkafmcomindexphppark-afm-modes41-

mediasresourcesbrochures (Accessed 11th February 2018)

19 Park Systems Corporation XE-120 High Accuracy Biological Sample SPM User

Manual (2013)

20 Castellanos-Gomez A Agraiumlt N amp Rubio-Bollinger G Dynamics of quartz

tuning fork force sensors used in scanning probe microscopy Nanotechnology 20

215502 (2009)

21 Crystran Quartz Crystal (SiO2) Optical Material Available at

httpswwwcrystrancoukoptical-materialsquartz-crystal-sio2 (Accessed 11th

February 2018)

22 Nakamatsu K-I Nagase M Namatsu H amp Matsui S Mechanical Characteristics

of Diamond-Like-Carbon Nanosprings Fabricated by Focused-Ion-Beam Chemical

Vapor Deposition Jpn J Appl Phys 44 L1228ndashL1230 (2005)

23 Yao N Focused Ion Beam Systems Basics and Applications (Cambridge

University Press 2007)

24 Deformation and height anomaly of soft surfaces studied with an AFM Precis Eng

16 68 (1994)

25 Sader J E Larson I Mulvaney P amp White L R Method for the calibration of

atomic force microscope cantilevers Rev Sci Instrum 66 3789ndash3798 (1995)

26 Weisenhorn A L Khorsandi M Kasas S Gotzos V amp -J Butt H Deformation

and height anomaly of soft surfaces studied with an AFM Nanotechnology 4 106ndash

113 (1993)

27 AFM Modes and Theory mdash Atomic Force Microscopy Explained - Nanosurf

Available at httpswwwnanosurfcomensupportafm-modes (Accessed 11th

February 2018)

28 [DI MultiMode Manual] Available at

httpswwwcigsunimoitCigsDownloadslabsAFM2manuali_lettureMultiMode_

Manual_RevBpdf (Accessed 11th February 2018)

29 Systems S R MODEL SR850 DSP Lock-In Amplifier (3ndash1) ndash (3ndash27) (Stanford

Research Systems 2009)

30 DeVore S Gauthier A Levy J amp Singh C Improving student understanding of

lock-in amplifiers Am J Phys 84 52ndash56 (2016)

31 Instruments D MultiMode SPM Instruction Manual Deflection Sensitivity 206

(Digital Instruments 2004)

32 Grober R D et al Fundamental limits to force detection using quartz tuning forks

Rev Sci Instrum 71 2776ndash2780 (2000)

33 Jeol Energy Table for EDS Analysis Jeolcom (2017) Available at

httpswwwunamurbeservicesmicroscopiesme-documentsEnergy-20table-20for-

20EDS-20analysis-1pdf (Accessed 2017)

34 Chatterjee N Electron Microprobe Analysis 6 (MIT Electron Microprobe Facility

35 Hull R Properties of Crystalline Silicon 137 (INSPEC The Institution of

Electrical Engineers 1999)

36 Zhang G Wei Z amp Ferrell R Elastic modulus and hardness of muscovite and

rectorite determined by nanoindentation Appl Clay Sci 43 271ndash281 (2009)

37 Khan S H Matei G Patil S amp Hoffmann P M Dynamic solidification in

nanoconfined water films Phys Rev Lett 105 106101 (2010)

38 Garciacutea R amp Paulo A S Attractive and repulsive tip-sample interaction regimes in

tapping-mode atomic force microscopy Phys Rev B Condens Matter Mater Phys

60 4961ndash4967 (1999)

39 Zitzler L Herminghaus S amp Mugele F Capillary forces in tapping mode atomic

force microscopy Phys Rev B Condens Matter Mater Phys 66 (2002)

40 Nagy P B amp Adler L Surface roughness induced attenuation of reflected and

transmitted ultrasonic waves J Acoust Soc Am 82 193ndash197 (1987)

41 Vorobiev A Gevorgian S Martirosyan N Loumlffler M amp Olsson E Intrinsically

tunable 067BiFeO3-033BaTiO3 thin film bulk acoustic wave resonators Appl

Phys Lett 101 232903 (2012)

Investigation of the Acoustic Response of a Confined Mesoscopic Water Film Utilizing a Combined Atomic Force Microscope and Shear Force Microscope Technique


Recommended Citation

tmp1532710449pdfV1num

Investigation of the Acoustic Response of a Confined Mesoscopic Water Film Utilizing a

Combined Atomic Force Microscope and Shear Force Microscope Technique

Monte Allen Kozell

A thesis submitted in partial fulfillment of the

requirements for the degree of

Master of Science

Physics

Thesis Committee

Andres La Rosa Chair

Peter Moeck

Erik Sagravenchez

Portland State University

Abstract

Acknowledgments

Table of Contents

List of Tables

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

Abstract

Acknowledgments

Table of Contents

List of Tables

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

Acknowledgments

Table of Contents

List of Tables

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

Table of Contents

List of Tables

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

List of Tables

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

List of Tables

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

List of Tables

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

List of Tables

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

List of Figures

sensor 7

tine 36

carbon tip 38

tine 40

tip 41

displacement 68

distance points 75

acoustic signal 92

acoustic signal 93

acoustic signal 94

(Image 1) 100

400mV 110

labeled 113

regions labeled 114

350mV 118

labeled 122

regions labeled 123

outlines 136

500nm 137

interactions

damage

2014 pp23-242

geometry

signal

distance (d)

Range (nm)

Intramolecular

(Ionic or Covalent)

1d lt1

Dipoles 1d3 05 to 3

Solvation e-d lt5

ion repulsion

experiment domain

Fsp = -2AR33z2(z+2R)2 (EQ 21)

Hj3 = 2724(HRtKc) (EQ 23)

zero-distance point

1 Nm 193 nm

10 Nm 089 nm

20 Nm 071 nm

30 Nm 062 nm

Spring Constant (K)

1 Nm 041 nm 1 Nm 416 nm

10 Nm 019 nm 10 Nm 193 nm

20 Nm 015 nm 20 Nm 153 nm

30 Nm 013 nm 30 Nm 134 nm

50 Nm 011 nm 50 Nm 113 nm

given as

following form

(MA) = (L1L2)

as a force sensor

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Estimated

Mechanical

Advantage

Applied

Cantilever

Displacement

Expected Laser

Displacement of

PSPD (um)

125 um 45 cm 3600 10nm 360 um

50 um 45 cm 900 10nm 90 um

1400 um 45 cm 3215 10nm 3215 nm

optical path length

L1 Cantilever

Length (um)

L2 Optical Path

Length (cm)

Mechanical

Advantage

Cantilever

Displacement

Displacement of

PSPD (um)

125 um 46 cm 3680 10nm 368 um

50 um 46 cm 920 10nm 92 um

1400 um 46 cm 3285 10nm 3285 nm

stiffness

Klat = tE4( w L )3 (EQ 25)

cantilever

Dimensions

Tip Radius

7667 um 3467 um 4854 um 577 um 124 nm

E(Perp) amp E(Para)

Shear Modulus G

E(perp) = 787 GPa20

E(para) = 982 GPa

G = 3114 GPa 21

IBD Carbon

Kver Klat Ktor

173 Nm 11019 Nm 17959 Nm

number NCH-W

Figure 34 Schematic

illustrates afm

cantilever transfer

process Image A

shows an unmodified

chip This is how it

arrives from the

manufacturer

Image B shows that

Image C shows that

fork tine

procedure

carbon tip labeled

Dimensions

Tip Radius

933 um 38 um 35-5 um 1500 um 110 nm

735-219 Nm

meaningful data

the cantilever

afm normal force

FN = -HR2(6Z2(R+Z)) (EQ 41)

of force Ft

Ft le 30pN

zero force

force is zero

Image from 27]

(119879119900119905119886119897 (119909 119910 119911) 119901119894119890119911119900 119889119894119904119901119897119886119888119890119898119890119899119905)(119899119906119898119887119890119903 119900119891 119863119860119862 119886119889119889119903119890119904119904119890119904) =

(119909 119910 119911) 119903119890119904119900119897119906119905119894119900119899

Calibration Profile

procedure

Figure 45 A high

resolution STEM

image of the low-z

measures 74nm in

height AFM

measurements

indicate a 75-100nm

perturbation

Τ = 1(2 π Fc) (EQ 47)

+-10 V

Figure 53 A laser

displacement curve

A sensitivity value

is the inverse of

slope which gives

Beta β = 2248

A = (2427)(KcRt) hj3 (EQ 51)

table 14 below

(red line)

60310-18 J (Zsnap = 257 nm)

60310-18 J (Zsnap = 271 nm)

Voltage

EDX Parameters

(Spot sampling)

Alpha0277keV

Alpha1793keV

Alpha9241keV

Alpha1098keV

Alpha1486keV

L-Alpha3443keV

Source33

K-Series

Atomic Gallium

KL-Series

Atomic Si

K-Series

Atomic Trace

Elements

EBD Thin Film

Carbon

(1 sigma 197)

NA 7114

(1 sigma 360)

Al-K-Series

(1 sigma 004)

Sn-L-Series

(1 sigma 008)

IBD Thin Film

Carbon

(1 sigma 437)

(1 sigma 216)

(1 sigma 121)

as ADC1

be needed

sections 42 and 43

between 55-65

50kΩ impedance

repeatability

Data Set 1

is repulsive

force bump

Data Set 2

surface

amplitude

14mv 14nm 288 umsec

25mv 25nm 514 umsec

W = 2πFResonance

Vel = -A0wsin(wt)

equation

this region

medium

AFM cantileverrdquo

dynamics

phase signals

are visible

not visible

red dotted outlines

Deposition

500x500nm2

Amplitude

LrEBD X X X X X X

surface analysis

unknown

References

(1998)

doi1015760etd2046

6081ndash6090 (1996)

Manual (2013)

215502 (2009)

February 2018)

16 68 (1994)

113 (1993)

February 2018)

60 4961ndash4967 (1999)

Phys Lett 101 232903 (2012)

Documents

Investigation of the Acoustic Response of a Confined